The same thing happened to me, but after some time and studying, I am understanding all these things. Now I am in the process of understanding the performance differences between different versions of Windows.
There is a windows update for the 9000 series. Its confirmed by AMD. They ran their tests on a specific Windows 11 version which can be downloaded, they posted an article on their site
And just before that Nvidia and GPUs in generally weren’t sold to gamers. Rather bots to farm fake money that didn’t go anywhere but into a few folks pockets and well of course the gpu manufacturers 😂
2008 intel silicon lottery to see if you can get your i7 up to 4, 5, or 6 ghz 2024 intel silicon lottery to see if you don't have to down spec your cpu to keep it alive
@@Amaeyth if you put it that way it sounds better than the only meaningful architectural jump since 2015. that's the problem here intel's 6th, 7th, 8th, 9th, and 10th "generations" are the same architecture on slightly better process nodes, with core and clock creeps. 11th gen was a lazy attempt at a slightly wider core from the previous years, backported to 14nm. 12th gen was intel's only jump in the ryzen era, and then 13th gen was a minor cache bump, and we're back to stagnation with the 14th gen. that's quite a terrible track record for the past eight years as far as innovation goes. six of these eight "generations" are just marketing, nothing more.
Blaming mobo manufacturers is wild. K, KF, and KS model CPUs are "unlocked" yet apparently any kind of overclock whatsoever just fries the CPU in a matter of months, while the current Intel stock values will likely fry it in under a year. We are going back in time when installing a component is no longer as simple as "plug n' play" like back when installing a new HDD required you to configure the switches on the HDD and go into the bios to tell it what switch it's set to, oh and also make sure is in the right position on the IDE cable as well. Except with my 14900KF, I need to update and reconfigure my BIOS each month, but hey, at least I don't require an Intel account to run my Intel CPU. Yet... We should at least get a 25% cashback considering the unlocked models are not safe to overclock, and as such, we paid a premium for a feature the CPU wasn't built to withstand, like a Wish product that only lasts a few months.
Hopefully US top notch lawyer's can bring a class action against Intel ( with EU smacking Intel a bit ) to offer all affected customers (of 13th - 14th Gen I7-19 Chips) - Equivalent of their New CPUs (15th Gen 1851 Socket Intel Arrow Lake ) and $200 - $500 for an equivalent Motherboard to house their new chip CPu. All this f^^&^&^& up is harming the Intel Brand and p^^%%^^% lot of us off; with view that our new build will unlikely be Intel but their main rival AMD; many say it's like the CPU equivalent to B loeing 737 max f&^^&^& up.
@@dafyddthomas7299 they probably won't, for two reasons: intel also has top notch american lawyers (arguably even more so because the us operates on the doctrine of the rich shall eat the weak) and because splitting intel apart and declaring bankruptcy on the cpu design business would probably be cheaper than providing all that stuff.
My 13900k is degraded. Can't pass 28000 on cinebench r23 with my Asus board on the new microcode. Even with extra power I can't pass 30k. Waiting on Intel with an rma
If your chip has degraded it will require more voltage for the same clocks. Basically the quality of the silicon will diminish and it will become less power efficient. Other damage from degrading could include instability so bad that voltage wont fix it, this is the most severe of damage and requires RMA. When you reduce power on a degraded CPU to curb the issue, it may only cause more crashing. This is due to the CPU now requiring more power than when you bought it. You will in most cases have to dial everything down but then up the vcore until it stops erroring / crashing. Disable multicore enhancements. Use the balanced power plan, enable C-states. Stick to a MAX of 1.35v vcore and monitor this with OCCT, HWINFO64, etc. Lock power limits to 253w for PL1 and PL2 and lock Amps to 307A or whatever you want below that. You can set adaptive voltage offsets with a minus symbol to further turn down the power your CPU draws, other than simply reducing vcore on its own. Another solution would be to lock your cores so that you disable the high power draw "6ghz boost on two cores" BS.
"There's a timer on your CPU and you should turn it off more!" Sorry, I'm used to running CPUs so long that they're hopelessly outdated before they run into degradation issues. I was still running a 5000+BE in 2016, and it still did everything it once did. Yes, it's electronics, yes, electronics have lifetimes, and yes, the 13th/14th gens are trying to kill themselves at what should be but isn't an idle state, but it's still quite shocking (pun intended) that we're now seeing stuff go to hell quickly enough that the replacements don't exist yet.
Same thing. Many old processors, even with OC lived for so long they simply became outdated, suitable for people who cant afford anything better. CPU dying that fast isnt norm, just as it isnt norm for GPU`s to fry themselves without doing something horribly wrong (like using molex adapter to 6, or, god forbid, 8 pin connector).
I'm a prime example of that as proof. I've been too damn lazy to upgarde my machine. My machine is a Core i7 965 first gen Nehalem. I've left this thing on all day for over 16 years. Granted, I never took major advantage of the unlocked multiplier or its overclock capabilities, but this thing has been running in a place that has climate very similar to Florida's (80-90F at worst, but high humidity) with moderate air cooling. And it still runs, it still doesn't BSOD, it still has no problems running the games it can actually run. Like you said it's depressing AF that my i7 965 from 16 years ago running all day in warm and not-so-ideal conditions is lasting longer than these 13/14th gen Intels. These things are literal ticking time bombs and Intel should be ashamed of their screw up.
its not only degradation. they had production issues. so most of the cpu affected are affected by this. above that probably a managment problem forcing the stuff to crank up the voltages and power consumption to insane levels to be able to keep up with amd in synthetic benchmarks. literally the first thing i did with my 13700 when it was new, is that i lowered the power limit in bios to 160 watts, and lowered the adaptive core voltage. my processor still beats prime 95 no problems.
I'm still using my Intel i7-2600 CPU @ 3.4GHz I purchased in 2013, which I still used today. The PC has been left on 24/7 all those years. The only upgrades have been an SSD and an extra 8GB of ram. I'm looking at building a 4K editing machine with an i7-14700K and a 4080, but the constant talk of stability issues is preventing me making the purchase. Maybe I need to get over the fact that it won't last another 11 years.
I have had my 13900K running on air since the day I got it (I know, I know). It had these same crashing issues since day one. So, from the very beginning my goal was to get acceptable temps. So I immediately did some BIOS tweaks, undervolted, lower maximum clocks, LLC adjustments, etc etc. With the tweaks I ended up making, my 13900K maxes out at 80C under full load on air. It also eliminated the crashing issues. The new BIOS runs perfectly for me without any tweaks, aside from temps, so I applied similar changes I did before to lower my temps and I'm pretty happy with the results. In a way, I feel like I, "saved," my CPU just by having it on air (against recommendations) because it required me to undervolt/underclock right away.
what is your cooler? maybe you need to replace cooler, if you have bad cooler, it may stop. i have 13700, and 14900. I replaced with highest cooler out there.
@@dxncx4555 You're missing out on 2.8ghz of performance though. Why not keep raising those clock speeds until your temps sit just under 80c while under stress
@@tilapiadave3234 Not quite the same. AMD's lies are about performance. Getting your AMD CPU and finding out it is slower than advertised is not the same as getting your Intel CPU and finding out it will be dead in three months......and that they might not replace it for you.
@@tilapiadave3234lie about what again? Is it the recent performance issues with the 9000 series? At least the problem was identified by reviewers before the launch, unlike Intel's issue, which took over a year to show signs of degradation after they sold a lot of their hardware to consumers, enterprises, and data centers. I would still prefer AMD's lies on performance charts over Intel's unfixable and irreversible damage, which they tried to sugarcoat you with a BIOS update. Intel knows that the issues lie within their hardware and not the BIOS. Yeah... Yikes.
Ugh. Got my 13700k in November of 2022 a few weeks after it came out. I was SO hyped to have a next gen CPU. My PC started having bluescreens about a year ago and they keep getting more and more frequent. BIOS updates, disabling XMP, reinstalling windows, nothing has worked. I am in the process of RMAing it now. Such a shame.
It’s crazy how you talked about Falcon NW right away because I had a 45 minute phone call with them where they walked me through the bios update and profile preferences. Thanks for promoting them on your channel, they’re awesome!
I bought my 13700k around launch back in 2022 and yeah the degradation started around 7 to 8 months in, started crashing while compiling shaders and such, specially on emulators. I RMA'd mine and to be fair the shipping and replacement arrived rather quickly, HOWEVER, be aware that in the email they send you with information about your RMA and about your replacement, it will state that the model they'll send you will vary depending on what's available, I got in return a 13700KF instead, I personally didn't mind much since I'll never use the integrated graphics and have spare GPUs in case of emergency, but keep that in mind if you do RMA your CPU.
@@1newme425 See, that's what I thought as well, and I agree, sadly I needed the CPU back as soon as possible as I had work needed to be done, don't have any other working PCs atm, loosing that part of the CPU was ok just to resume my work. That's why I posted this so people are aware and call intel out if they notice a smilar RMA.
Many people who RMA’ed their i9-13900k’s and got an upgrade to i9-14900k’s. Meanwhile they technically downgrade you. They gave you a product lesser than the one you had originally. That’s absurd.
Don’t know why people still listen to jays hardware advice. He’s a cool guy, but he just has no idea what he’s talking about when it comes to this sort of thing.
I know right that was bothering me the whole video. Then again maybe he's smarter than we think and he's aligned his loadlines just right so they are the same. But probably not.... EDIT: People seem to not be reading the last sentence of my post. I don't think this is the case, which is what "but probably not" is supposed to mean.
Step 1: low gpu usage beyond normal? Step 2: random lock ups even at idle Step 3: USB flash drives randomly go offline or not recognized Step 4: Any bluescreen? Step 5: heavy crashing Unreal 4 or 5 Engines Step 6: Out of Video Memory error when starting Red D Redemption 2 Step 7: Random restarts without any overheating or lock up This is what I experienced on both of my failed 13900k cpus. That was also undervolted, watt limited, and cores locked Never drew more than 1.35volts in any scenario. CPUs began showing 1 or 2 symptoms above and eventually all of them. Was happening at 6 months use marker on both cpus. It is a shame as 13900k is amazing for workstation and gaming, but I am already tech level gut that is dealing with isolating issues with cpu. It happens again at 6 months marker. This repeatability is an identifier of identical issues within the cpus' designs. Same failure points. I never encountered same identical issues in 25 years of building pcs, down to amount of months
@@SwingArmCity I was being feciecous...fukin around... I actually have a an i7-133900-KF maybe misplaced a 3 or 9 (or whatever) basically the processor KF vs just the K version is supposedly completely unlocked with no restrictions on it... it's been a while since I bought my latest pc...so forgive me if im not 100% accurate... But these damn octocore 12th gen processors sometimes acheive almost 5k htz range...which is fucking insane...i started underclocking for the sake of how damn hot it got in my room..lol I even removed the glass plate that is the removable side of the case in case you got to get inside for anything and replaced it with some very cool perforated screen specifically to protect against dust, but let's the entire pc breathe and disperse enormous amounts of heat to which I added a mini tower fan about 1 ft tall that helps ventilate the pc as well as the area around it... Much better than liquid cooled stuff in my opinion... No matter how hard i run my pc (also has a RTX 3060 ti (the bad ass kind ppl use for mining bitcoin), plus the 1000 watt power supply.... ...the processors or GPU never get above 80 degrees centigrade....which most of the time operates in the 60-70 degrees range on tha reg... Which accomplishes (1) The sheer heat the entire rig typically gives off.. (2) allows the PC to run as hard as i want it to without overheating...(3) helps stave off degregation for much longer than running sealed in it's fusion level hot generator sealed metal box compartment.... Anything electronics that run with capability and intention for performance will ALWAYS break down over time....it's physics...💯 THERMODYNAMICS ARE VIRTUALLY IMPOSSIBLE TO CIRCUMVENT.... EVERYTHING WILL DEGREGATE or deteriorate over time.... But the only way to extend the life of your performance based electronics (specifically PCs) you just have to get creative or specific with the hardware you decide on to get the most viable life out of it
Had my 13900k for almost 2 years no blue screens, freezing or crashing. Usually upgrade every 2 years anyway so I'm not going to lose any sleep over it.
Like you I am on a 2 year cycle as well, but on a 1 generation behind schedule. Meaning I dont go with the newest generation when I upgrade I go one generation behind, that way I miss all the growing pains of that newest generation.
@@KingDooksare you thermal throttling? My 13900k actually gained about 1500 points after undervolting. R23 score was around 34k, now it's passed 36k. It somehow runs at higher clocks with less voltage. And I'm also not using an unlimited power limit which is how most people get really high scores.
I am really surprised by my midrange Asrock board. First vendor that did not pump insane voltages on auto settings. Asrock is by far the best board manufacturer. Okay maybe I got a 1% CPU with an insanely low VIP table, but Asrock truly rocks. I am an Asrock fanboy now.
Electronic components initally powering on is hardest most stressful thing it can do. This is why when a light bulb bursts, its typically as you flick the light switch (though now-a-days LED bulbs are a bit different but same principle applies). So personally I wouldn't adhere to that advice at the end about turning off your PC all the time that you're away from it. You'll potentially cause your CPU to degrade faster, along with every other component in your PC, depending on how often you actually use your computer. As someone who uses their computer every day, i almost never shut off my PC. I have an i7-950 from yesteryear that still functions today, and was something I pretty much ran 24/7 for nearly a decade.
I have an i9-13900. I never had any issues with stability or crashing, but I upgraded my bios to the latest to ensure I had the new Intel microcode. For the first time ever, Overwatch crashed on me yesterday. I'm concerned the new BIOS is actually causing me issues, but I only have the single crash so far.
that means you processor has degraded, you will need to add a low off set and lock the cores, the reason why the new microcode is crashing it is because its too late and the undervolting is triggering the crash.
I’ve had the i7-13700F for about 7 months now and have zero issues whatsoever with it and been hesitant to update my bio for the new version for that reason as I’m not sure if I should really do it if I’m not having issues
@@bardbacx I suspect the BIOS was rushed out and has bugs. If the CPU was degraded I would have expected stability issues prior to the "fix"". Thanks for the input though.
Built a PC less than a week ago. MSI MAG B760M WiFi II motherboard, gen 12 12700k, 64 GB Corsair Vengeance RAM, MSI RTX 4060, Corsair SF850L PSU, MSI 240 AIO, MSI 500 M2, 2TB 2.5 platter backup drive crammed in to a Jonsbo Z20 case. Yep, I had RAM OC errors before BIOS update. Not currently. Runs smoothly, for now. Love it! I'm not an overclocker or heavy gamer. Videos like yours kept me motivated.
My blue screens were caused by my boot drive getting too hot. It took a long time for me to figure that out. It started the moment I swapped to the 13600K from the 5600g. It’s definitely the difference in power draw, so the 13600K ran hotter. I fixed my problem after swapping to a new case.
I just switched my i9 14900k for an i9 12900ks and I'm loving it. I'll probably switch to AMD next build but the 12900ks runs like an absolute dream and is a beast. Such a shame that Intel have blundered this so damn hard.
I happened to switch right before all this went down. Got lucky I suppose. If you do, dont be like me and forget you have to flash amd bios to get it running. Man I spend way too long troubleshooting nothing.
@@Stony1313 There is a list of steps for EVERY CPU and new motherboard. (And of course, one for EVERY used CPU/Motherboard combination.) If you stick to a plan every time you get a "new" system, you never waste time. :) I know it takes (upfront) time, but it saves hours troubleshooting.
Yeah I switched to team Red for the first time in my life, right before this crisis hit! For once in my life, something worked in my benefit! 😁 In comparison, I once bought a v8 Chevy. The next day America announced sanctions with Russia due to Ukranian war. Making gasoline skyrocket in price! 🤦♂
I have an I9 13900k I've been using since May of 2023. Not sure if it is degraded or not but I really don't want to do all those tests to find out. If it craps out one day I'll just replace it, no big deal.
If you get any out of video memory errors, it's your cpu. Had my 13900k for about the same time and had these errors start about a month ago. Now it's fully bricked. Good luck!
Dear Mr. JayzTwoCents, I just want to say thank you for all you do. Your videos have been very helpful when I'm trying to figure out something with my PC.
I'm still on the fence about upgrading the BIOS for my 13600K on a Gigabyte motherboard. It's super stable, runs under 1.25v max with temps at ~72C while gaming. I've seen only 2 games that push the power over 100W (excluding shaders compilations).
13600k here too. Not seeing any issues, not going to update to new BIOS because of performance issues. If it dies I'll go to Intel I guess. I got mine in 2022 though. It was speculated that the original first batch don't have this issue. Has anyone else heard that?
im on Bios version 2305 with ASUS z690-F and I have spent some hours some days and cant see any voltage problems or anything, I have HWmonitor and tested with Cinebch and voltage dont go over 1.251. I do XMP 2 and only lower offset voltage, never touched LLc or anything else, I run my computer stable, never crashes im on 13600KF
And most of them will get there cpu sent straight back and charged a disgnostic fee. Be damn sure you need rma before you kick and scream over the phone.
FYI, I have been having constant crashes when running Diablo 4. I tried running Starfield and it Crashes.I tested my new NZXT-PC with a i9-14900KF and both Cinebenchs used in your video and they crash and OCCT wouldn't even start, and when I ran Intel's Diagnostic (latest version from Intel), the Burn Test, Unigen and NVIDIA install had no issues. I swapped out my Gigabyte RTX4090 for my ASUS Strix RTX2080Ti and no crashes. Clearly I am having video issues. Thanks for the great video in helping me identify my GPU issue. Stay strong and thanks for the great instructional video.
the only intel I have is at work. It's a dell. It's in the affected list. I'm hoping that dell being crap means that it was never hit hard enough to worry about. It took forever to upgrade from my 4th gen CPU. They really want laptops here. I HATE laptops.
It's long been understood that 70 degrees C is is safe sustained operating temperature to prevent circuit degradation. Never had a CPU fail for over 10 years use 24/7 adhering to that. Aim to keep the top temps below 80 degrees C under load.
1st thing I did when I got my CPU was undervoltage it to 1.2V. It has been super stable and the temps are great. Stock voltage in these generation is crazy, temps are crazy, everything is kind of insane. I guess is all for the bench wars.
I just under volted to 1.250 and it went back up to 1.450 when Windows loaded. I think I’m just going to do an off set in the bios to get it below 1.350. XTU doesn’t allow me to undervolt since this new bios update. 14900k.
The problem is, that Intel should set the stock voltage at healthy levels (and clockspeed at stable level accordingly). And let people themselves take responsibility if they want to OC / OV the CPUs. Intel didn't really needed to be Faster than AMD. (They already were the most popular) They got that succes by being the most stable and easy to use platform. They managed to survive and even grow with the dreadful "slow but stable" Pentium 4 against the faster Athlon for years.
Very good video. I built 3 custom systems spread over the last 18-20 months for a local business. A 12th gen and 2 13th gens (all i9's, high end builds). The last 13th gen i9 failed after 2 months of use (at a business used 5 days per week, min 8hrs per day). This happened before it became widely known about the issues with Intel. I RMA'd the motherboard and CPU just to be safe, and I didn't know which component caused the failure (to simply turn on). BTW, the replacements worked. I have told the business to check their MB settings and update the BIOS on those boards.on several occasions. I don't think their IT person has done either. I built myself an i5 13th gen and have set the motherboard to the correct default setting, for a while now. All thanks to your videos and information.
I was offered the same they then came back to me to say they have some new CPU's available. Im still requesting the refund (Told its going through the approval process in upper management) I will be going with the 12900K.
… Brilliant … thanks for going through these tests with us … as ever you are responding to the needs of your community👏
2 месяца назад+18
At no point did you check the actual realtime voltage until you glazed over it in OCCT. Pretty sure MSI boards use die sense so you should be able to see the voltage the CPU is actually receiving. The VID and the actual Vcore are not one and the same. Also you neglected to mention that current draw is the main factor in electromigration, and not voltage. Also the instability in shader compilation people have seeing is very much an all core, current intensive workload. A lot of misconceptions here that aren't quite right.
Thanks for explaining this with sentences. Others were giving 1 liners on the subject, short sentences and not saying enough to be worth my time to glance over their comments! Everyone upvote useful comments like this one please!
2 месяца назад
@@yonghominale8884 The MSI comment is in regards to die sense measurement using differential sensing pins being used and not meaning to say it's only applicable to MSI boards. Yes, however simply saying "it's ohm's law" doesn't help this video as it's simply saying voltage is the primary cause. We're not even mentioning the impact CPU load load has, or the fact that different loads cause different current patterns.
Had same issues you spoke about for my 14900k during first 3 months of 14th gen release. z790 hero / z790 strix e Artic liquid freezer II 360 Used two different motherboards and two different 14900k chips. Both scenarios produced the same results. Cinebench and games would crash upon starting and during cpu high spikes. Everything was tested stock settings out the box. Returned both chips and went with a 13700k. Everything has been running stable since.
Thanks JayZ because before this whole incident happened you post a video about motherboard manufacturer made a mistake on having a higher power profile than intel intended also provided us a guide to undervolt to mitigate the instability issue that's why I able to dodge the bullet on damaging my CPU.
@@kriszhao80 Ofc the fault is on both of them if motherboard manufacturers didn't put the unlimited wattage the volts will not go higher. And it will not trigger the problem.
@pacspacs8693 Intel literally sells unlocked processors as a selling point saying that they are designed to be overclocked and has been using the overlclocked benchmarks for years to try to be competitive, you can't promote the behavior then when you make a faulty product blame it on the motherboards who you have been encouraging for years.
@@kriszhao80 Exactly, and yet Jay goes out of his way every time he talks about this issue to throw shade at the MB partners. None of this is their fault. Intel just looked to shift the blame before finally admitting that it was their borked microcode that was 100% at fault for these problems.
Hi Jay, great work as always. As an RF, Analog, Mixed-Signal CMOS integrated circuit (IC) design and layout Engineer, I can tell you that, in general, the Silicon Chips, including cell-phone, desktop and laptop communication chips and processors are all designed for a 10 year lifespan with the highest currents and the resulting electromigration. Some last longer, some last shorter, depending on the process nodes. The computer I'm using to write this message is a 14 year old HP Pavilion with Intel's original i7 860 processor with NVIDIA GeForce GT 730 (that was installed later to allow for 4K display - since I'm using my 55" UHD TV as my monitor). I do turn off this computer between 8 to 16 hours a day. The first and only gaming computer that I built with your help and advice (thank you Jay - and Linus Tech Tips) uses i9 13900K CPU (liquid cooled) and NVIDIA 4090 MSI SUPRIM liquid cooled GPU. We'll have to see how long that will last.
Built in Feb 2022, my 13900k/4090 has not degraded. But only because I believe I took the precautions at the time an extra step than even content creators in that era had recommended. They called for PL1=PL2 limits of 253W. I set mine to 125W instead. And ran the 4090 with power limits at 70% in Afterburner. With enhanced turbo off and a -0.08V undervolt. My friends were curious as to why I chose to run everything so low. At the time, it was simply because the heat of sitting next to my pc was uncomfortable. So in the 80C and even higher 70C range, that was already too much for me. My settings allowed me to use my computer at 50C and 60C instead. At the cost of some performance.
Kinda sad though... investing so heavy in a system to then downclock it, but hey happy that your CPU is doing fine! :) My 13700k is not experiencing any issues either
@@squallsquall9669 Some don't care about getting every little bit of speed (but close to it), some just want stability and enough speed close to what was advertised to play all the games and do all the things they need and just have it work. Also some when they bought the processor did not know they would have to do all this undervolting, changing of settings etc.. that came to light. Because all this did not come out till later.
I also have my 14700KF undervolted at -0.08v. Set PL1 135 and PL2 150 based on observation during gameplay and loading screens (time to compile shaders, etc.) The games I play typically use about 175w to either compile shaders or load from area to area, so I capped it at 150w peak for 56s and don't even notice a difference (probably milliseconds in loading). Same during gameplay, which peaks around 130w, so I capped it at 135w so it doesn't hinder performance while the undervolt keeps things cool. I generally don't see anything over 60C in the most intense scenarios (generally low 50C) during gaming while fans are set around 50%. Idles at around 30C. Volts usually peak around 1.35v at 5.5x ghz during intense gaming. If I ever need more temporary power, I'd just up the PL2 back to 253w worst case, but I'm in a good place atm.
@squallsquall9669 believe it or not, undervolting and limiting the watts only really affects benchmarking and has little to no negative drawbacks during gaming (not sure about rendering etc.). In real world gaming scenarios, dropping temps down 20C while maintaining near spot-on performance from stock settings is great. I notice absolutely no difference during gameplay and load times. It could be because I game at 4k so everything is more gpu reliant, but hey, it works perfectly as intended.
If love to see a more detailed vid with OCCT, either OC’ing, stability testing, undercoating, memory OC’ing or something to explore the tool more, there’s of options. Love the info as always Jay!
I personally would happily pay for OCCT except they want to make it a subscription. Being a subscription offends me, because it's designed to hope you forget to cancel it. And this is just for small businesses/home users. Not to mention, it does *not* include all of their enterprise commercial license agreement features. I asked the owner about that because as a paid home user I wanted to use those features, and they are not going to bring them for anyone other than enterprise.
Lot of people complain about MSI's quality, but we have this happy coincidence where MSI using lower power limits (possibly to compensate for lower quality parts) has spared their 4090s from melting connectors and now possibly spared Intel CPUs from early degradation.
@@johnt.848 I didn't say it did. I was commenting on how two completely different issues were _coincidentally avoided_ by having lower power limits and the reason MSI has lower power limits on both their GPUs and mobos is probably because they use cheap parts. It's an amusing irony the brand with a deservedly mid reputation is doing better than everyone else because of this.
MSI's 4090s are also burning. The issue is due to the connector not having a tight enough fit in which a slight gap from the cable being bent too much causes arcing, which produces enough heat to melt the connector.
@@Skelath Unless there's been new information out, from what I recall in the North Ridge and Jay's videos is that MSI is underrepresented in the available data regarding melted connectors. The only way that's possible is if a) nobody's buying MSI (unlikely, they would've disappeared before EVGA if that were the case), b) they have such an amazing RMA that nobody's sending them into third party repair shops (also unlikely, because MSI's RMA process is not amazing), or c) something about their cards coincidentally avoids the design flaw of the new connector.
Lock all the cores first at the real speed of the cpu, for example the 13900ks is 56 (5.6 ghz). Then after, set voltage to 1.345v and a negative offset starting at -75, and work you way up if you need more of an offset, you want to keep the "vid max voltage" below 1.35v as anything above is too much. Also, power limit the cpu to 253 watts max will also help and make sure if you have Windows 11 that you turn off "core isolation" as if you don't it overrides any voltage settings, and offsets you set. This is what has worked for me with my now 2nd 13900ks and been going good so far. Hoped this helped!
@@rleekc Ulike the original comment here, it has NOTHING to do with your CPU. This is simply the DisplayPort doing what it is supposed to be doing. DisplayPort essentually conducts syncing operations with the monitor, it's called link training. This link training can happen at any time and it is done so in order to maintain functionality of the DisplayPort link; what it is doing is reducing the link speed or increasing the link speed, which is based upon the best link speed while removing errors. Again, this is the result of a lesser quality cable if it is a regular thing (once a day perhaps). However, this is nothing to worry about at all, you can either ignore it or get a better quality DisplayPort cable.
Same, i assumed it's a video setting or something. @@rleekc I will largely get blank screens when trying VR, and toggling between apps.some windows it wont like.
Great video Jay. I have a 14900K on a Gigabyte Aorus Z790 Pro X board with a 4080 and 64GB of DDR5 6000mHz Ram. I was getting the “out of video memory” errors on high shader games like Hogwarts Legacy and Horizon Forbidden West and I do video editing on Adobe Premier Pro and I could not export a video no matter how big or small my CPU would thermal throttle and crash to desktop every time I tried to export a video. Since the BIOS update with Intel Extreme Default Profile and the new Microcode, I can load and play Hogwarts and Horizon completely fine now, no more “out of video memory” errors or crashing and I can go into Adobe Premiere Pro and export all videos now, my system is so quiet and exports them successfully. The crashes I was experiencing was happening EVERY time, but since the BIOS updates, it doesn’t do it anymore so I would so based on my tests my CPU is stable and has not “advanced” degradation, these are the tests I did and I believe my system to be stable now based on my experiences.
Every CPU is degrading, and will fail one day, as the earth will stop spinning one day, and the sun will die. With 13th and 14th gen we are talking about EXTREME degradation. I have a colleague (asus z790 strix MB) whose 13th and 14th gen CPUs died after 3-4 months of normal usage. And I have i7@920 that I purchased in 2008 and gifted to a friend years agoand he is still using it. So let's not try to intentionally or not diminish how BAD the situation is. And yes, I was an intel fan currently owning 13900k with MSI z690 unify. No problems so far for 1+ years, but after I installed the latest bios that was released today, cinebench R23 dropped from 40800 to 36500. All settings always defaults, no OC what so ever, not even XPM, because I'm always counting on the reliability instead of performance. And was willing to pay for a new motherboard every 2 cycles because "Intel is the best". No more Intel for me.
"Pimpin' been around the since world started turning. And it's gonna keep right on turning right along with it. Until this little planet rotates off it's axis as a result of it's core overheating and explodes into cosmic dust." - Black Dynamite
I needed this video! Should take a load of stress off my shoulders after the scare that my GPU pulled on me (no image output) after the 0x129 BIOS update! Thank god for the win+ctrl+shift+B shortcut to reset my graphics drivers without the ability to see my screen
my 4790k ran like a champ for 10 years now? i dont remember shuting it down more than once or twice a year. never thought of "shut down your computer" when you dont use it. now with my 14900k i might start doing that lol
I never once thought "I'm gonna be away from my pc for more than half an hour, and have no need to leave it on, so better leave it running for absolutely no reason".
got a new 13700k , updated bios and changed lite load and put voltage offset and now my voltage is at 1.250 while at load and its super stable and cool. Was a big upgrade from my 10700k in gaming.
@@therealsentinal4519 Go to mobo manufactures page , for my mobo it was MSI Pro z790 A Max and check under support / download you'll find the bios and " click here for a video on how to update bios " msi made a video by themselves showing all steps , just follow the steps and it's done After bios update only things I did was - 1. Turn off multicore enhancement ( called enhanced turbo on MSI ) 2. Change cpu lite load to whatever number you want , mine was at 20 as default so i brought it to 10 3. If everything is stable after changing above stuff run some tests or games you play regularly and check temperatures, performance and voltage and then you can start applying negative voltage offset in bios slowly lowering voltage till you find the perfect spot for your CPU and you're done.
I have had a 13th gen for a while and in the last ~5 months been experiencing random crashes. I just learned about the voltage issues a week or so ago and have been wondering exactly this. Is my CPU too far gone? Thank you so much for this video!
First Descendant shader compiling during start-up is a good stress test too. If it fails and the game crashes, you got problems. If it survives and the game launches, you're good.
Just upgrading the bios can be risky if you are on the MSI board. The first thing you have to do after that is checking AC/DC loadline either in HWinfo or in the bios. It shouldn't be higher than 1,1 mOhm. But MSI is pushing 1,7 mOhom for most of their boards recently which causes EXTREME overvolting. The easiest noob friendly fix: go to the BIOS and set Lite load mode to 12 (it should be at 0,8/0,8 mOhm). This way you don't have to change LLC from auto and keep CEP enabled without loosing performance. You can also change AC/DC LL manually but I don't recommend messing with that if you don't know what you are doing.
1:40 this is not entirely accurate information. YES Chips do degrade, however under 'normal' use that degradation takes HUNDREDS of years. Your CPU and GPU will live FAR longer than YOU ever will. It is only under excessive heat and voltage that the degradation is accelerated. With proper cooling your components will last literally forever.
More like 5 to 10 years for cpu of this century. That you double if you use you cpu only during the day. Litographie effect can reverse easily in super thin transistor
Yeah, but like it's fair to say "it doesn't degrade under normal circumstances" cuz the process is too slow to take into account for an user who'll use a cpu for like 5-10 years
My 13900K was bad from day 1. I must have gotten one of the oxidation ones as I got mine in Nov 2022. Consistently crashes when compiling shaders for certain games. Particularly those using UE. Otherwise, it works fine for older games that don't use 100% GPU. Starfield, though published after CPU release, uses an older engine, which works fine. Fails DLSS test on 3d Mark. RMA'd 4090 twice. Thankfully, you figured out it was the CPU. Had no trouble getting RMA as I had a boxed CPU since I built PC from scratch with parts from Amazon. And your build videos to put it together. Waiting for new one to come tomorrow as I got advanced RMA to eliminate downtime.
Thanks for tackling this, Hardware Unboxed and others won't touch the microcode updates or any of this stuff. They say their not AMD fanboys, but they won't lift a finger to support Intel users, so, I think they are.
Bro, i dont think these channels (maybe except linus) do that. Gamers nexus is waiting on the failure analysis lab reports and has mentioned they have a multi part intel coverage coming. HUB has been calling these amd cpus a waste, and made a 30 min video of amd shooting itself in the foot with bad products etc. Comments, on the other hand... i would 100% agree. Its ridiculous how much fanboyism there is
@@ZenAndPsychedelicHealingCenter He said it was a series of tests to " Hopefully " determine the state of your CPU. The video was more informative than what you posted.
With my 13900k and 14900k, I have never been able to run Cinebench R23 multicore without several cores hitting 100C. This is with an Arctic Freezer II 420 AIO with 6 fans in push/pull orientation. I am blown away by the temp you are showing there. I've tried remounting and repasting, worried it was a contact issue, but nothing has changed. Both boards I am running are MSI. One is Z790 MEG Edge WiFi with DDR5, and the other is an MSI Z790 Gaming WiFi.
I got some friends looking for a pc and they keep sending me intel builds, I cant recommend anyone getting an intel cpu for a while till i'm sure it's ok.
Jay!! I want to leave here my feedback! Running a 13900KF, on a Strix Z690-G. After microcode 0x129, I was getting max VID at 1.50 to 1.52v with Intel Default Profile, MCE off, Intel Fail Safe VID bahavior and enforced power limits, getting 35K in R23. I have tried the Asus profile, with MCE on, Auto VID behavior and disabled power limits and for my surprise, im getting 1.42v max 😱 and better Cinebench R23 scores at around 40K. No manually tuning. Never had any issues with my cpu in almost 2 years of use, but got worried about this voltages and degradation issue and updated to this microcode
Hey Jay, great video as always! However, I noticed that you’re using VID to show the voltage sent to the CPU. Am I wrong, or shouldn’t you be checking the Vcore value, which is listed a few lines below in the HWinfo? From what I understand, VID represents the voltage requested by the CPU, while Vcore is the actual voltage delivered to the CPU by the voltage regulator, and these values aren’t always the same. Please correct me if I’m wrong!
Think Jay needs to do some more research on this topic. I have a Asus Z690 and Asus profile undevolts my 13700k at base speeds 5.3mhz. The default Intel profile OVER VOLTS. This cpu will overheat nearly instantly on the intel profile and can barely hold 200 watts because it's running at 1.55 volts. The Asus profile under volts at 1.45 and can sustain 253 watts on any benchmark. I consistently get ~30500 pts on R23 with the Asus profile, while the intel can't even muster ~27000 because it's overheating so badly. Not sure what processor Jay is seeing Asus is overvolting on as the default Intel labeled settings are what are over volting. I'm personally not having any issues, tho I know the 13700k is not as affected by this issue, but it seems some are. I just want to know what CPU and MB Jay is talking about instead of just blanket calling out a brand. As I wonder if he understands which combos he's referring to. And yes this is on the most recent bios 3802. But this behavior is the same as its always been, even before the microcode update as I got mine in Mar 2023 and it's functioned this way since the start.
The safe voltage is 1.3 but it reduces performance. Intel won't force you to get to these levels otherwise they will have falsely advertised. Instead the microcode just prolongs the CPU enough to get past warranty period. Thought it was well known the P cores are the problem.
@@McFamily86 that's good to hear. You'll might crack 30k by making sure nothing is running in the background when you test it or changing the priority to high in task manager. But high 29k score should mean you're likely just barely over saturating your cooling. I'm using the Arctic 280 AIO. There's so much confusion with this topic. Sadly Jay is adding to it imo.
@@glyntaylor-williams3593 Yea this is not a MB problem in my opinion. Sure they didn't help. But Intel is straight up lying about the performance and volt requirements. I've even gone as far as to just set my 13700 to the 90 degree max. I only expect to keep it for 5ish yrs as it should last that long, and be powerful enough to meet my needs for that time period. When I bought it, I had all the expectations in the world it would easily last that long and that cooling would not be a major issue. Boy was I wrong. So far I have no crashing issues, I did at the start of ownership of it however. As figuring out the correct voltages was a nightmare because everyone's' "use this xyz" setting did not work. The variation on the 13/14 gen in what works is wild. Which really re-enforces the it's not the MB's causing these problems. It's Intel. Final thought, I also have 2 P cores that LOVE to blast right to 100c on the simplest of workloads. They are not the prefered cores either. Core 3 and 7 for me love to light themselves on fire at the first chance they get. This does worry me about my chips long term survivability. Time will tell. Likely when it's out of warranty....
I finally found my answer and to this day i have 0 issues with my chip! 13500 is based on the Alder Lake die. It should not suffer from degradation issues.
I know it's cool to hate on ASUS right now, but ASUS along with Gigabyte both significantly undervolt the CPU's with their default settings vs Intel settings due to the setting of the AC LL set to 0.4 mohms. In my testing, the "default Intel settings" is what overvolts my CPU the most out of anything right now. In both ASUS and Gigabyte BIOSes, you can manually set a voltage limit so you can stop it from going over whatever voltage you consider safe. I personally don't want mine hitting 1.55v which is Intel's limit with the new microcode, so I set my max to 1.4v or 1400mv. MSI refuses to add this option to their boards. Bottom line, if you know what you are doing, NOT using the Intel Baseline Profile can be better than using it.
Man I feel better or a little more at ease now as my 13900KS. It performed pretty close to your CPU. I was not having the problems that everyone else was having and did not experience any crashes before all this hoopla so I think I got really lucky with my CPU (knock on wood, lol). I ran the Heaven, the R23 and I have a Nvidia RTX 4070ti and always install the newest drivers and have never crashed on installing those, so I think I'm good. (knock on wood again, lol). I also have an MSI Z790 MEG ACE MB with the latest BIOS and ME so thank you for doing all this with an MSI BIOS and not ASUS. Thanks Jay for making this Video!👍
I wonder if the Intel problem has to do with the chips bending, because I have a contact plate installed on my 13900K and have never seen any of these issues.
@@PreemzI had a ThermalRight contact frame installed and recently a lot of tweaking undervolting and underclocking made my CPU displaying BSODs when running cinebench R23 then I started trouble shooting by removing Ram 1 by 1. The test was passing without bsod with only 1 ram installed then there was last hope I removed the contact frame and I ran the test several times and also stressed for 12 mis. No sign of BSOD. I don't know what I did wrong whether I overtightened the screws or may not tighten enough but I suggest you try it with Intel's stock ILM maybe you don't get BSOD again and do let me know.
@@gameotic1I was very careful to not overtighten the contact frame when first installing, and was very careful to apply even pressure across the IHS with quarter turns. Anyway, I've spent enough time troubleshooting Intel's mistakes already. I've also changed platforms over to AMD this past weekend, and am in the middle of refunding the chip. I'll sell the rest of the platform locally and just eat whatever the remaining cost is.
@@Preemz I've watched Gamers Nexus and other tech tubers guide and installed two of contact frames one for my own system and one for brother's both were working fine until last week my setup after lot of tweaking shown up this issue but my brother's system is still running fine. Maybe I haven't tightened the screws for my system and this problem emerged but I don't think it's necessary because a lot of people are using stock ILM and it won't bend the CPU really easily. Also with Intel's stock ILM the thermal paste distribution was very even so I don't think I need that thermal right contact frame again...
I have MSI MEG Z790 ACE mb with 13900k from 2022 week 46 and its been rock solid. I didn't even bother to turn off XMP and it passes IBT single and all cores. When it dies I'll get something else but it is under warranty now until 2028 per the intel warranty checker. The 2nd MSI 0x129 bios 1.C0 won't allow the CPU core voltage to run up like the previous ones did. FWIW Intel based systems have always been more stable for me than AMD and I've been building my own for over 30 years.
This is why I miss Intel OEM boards. You wouldn't have this problem with one that is calibrated for the processor properly in the first place and not monkied with by default.
Intel definitely have reference boards, they use them for internal V&V / release testing. I am super curious (1) why Intel does not sell them like they used to, and (2) why the internal testing did not catch any of these issues. Voltage spikes were clearly seen, even if they are on a non-degraded CPU
Ok someone feel free to correct me here but heres my ideology. What I think a intel microcode is like the cpu's own "default bios settings" that it will solely abide by. Then theres the UEFI where it intefaces with the cpu to deliver correct power and do all ur overclocking and stuffs. What I assume is that ur UEFI has way more power than ur CPU's Microcode therefore if you manually tune everything to low voltages in the UEFI, it would override the microcode "default bios settings". So Im currently thinking I dont need to update my bios at all since I already tuned my cpu in the UEFI to always consume 1.25v no more no less with stock clocks and hyperthreadding disabled. There has been no degredation and I haven't been getting any WHEA errors in event viewer. Someone feel free to correct me here!!!!!!!!!!!
Jay being real and helpful. I’ve noticed some other channels who were screaming about this problem have been silent since the microcode update. They slammed intel as the sole problem and said the chip’s performance would have to come way down to fix the problem. So now we know: -Intel was at fault. -The board partners were also pushing the cpu like they shouldn’t have been. -Intel should replace bad chips and extended the warranty. -the micro code doesn’t degrade performance much if at all. -Jay has been the most measured on this (IMO). He has been able to do testing, etc, but from many others not a peep. Is that because there’s not much new to scream about any more ? Btw I run a 5800x3d not intel.
Degradation has always been a thing if you pushed your CPU hard enough: My 2600K from 2011, could original do: 5050MHz without HT 4850MHz with HT 4 years later, 2015, it had degraded to: 4850MHz without HT 4700MHz with HT 8 years later, 2019, it had degarded to: 4700MHz without HT 4600Mhz with HT 11 years, 2022, I gave it to my nephew and it has degraded to: 4600MHz without HT, which is how he has been using it for a couple of years now 4500Mhz with HT
And my 2600 K is running fine. The issue is you are running it outside its design spec. If you didn't, it wouldn't have degraded. It's still running well, well beyond its design limits. CPUs will not degrade if run within their design specs - within any reasonable limits that a normal user will see. What kind of voltages were you giving it to get that 2600K running at 5ghz? Just like the issue with 13/14th gen, you were running it beyond its voltage design limits. Then it will degrade. Jay's allegation that there's a "ticking clock" is not true, unless you're pushing it well outside its envelope. You/ve got a ~135% overclock, even after 11 years of abuse. The fact that 13th/14th gens were degrading in *months* is another story entirely. Also, your degradation is -likely- not just the CPU. Your electrolytic caps on your motherboard are drying out, and the board is becoming noisier. I would be surprised if a re-cap didn't bring you back up somewhere closer to where it was new. Capacitors and other passives on the board are way, way likely to "degrade" well before any CPU.
@@christodd3361 You are right and it still running stable at 4600Mhz without HT, so not complaining 🙂I always ran it 100-150Mhz below the maximum OC and never had any stability problems with it. I have not consider that it could be the motherboard caps and it has been running in the same ASUS P8Z68 DELUXE motherboard for 12+ years, so maybe it is the motherboard and not the CPU. For voltage I used voltage offset for the CPU and believe the maximum voltage was 1.45 V, although it rarely hit that in games.
To say check if your cpu is ready for RMA is carzy!! Not just CPUs but any and all new hardware should outlast all upgrade paths and replacements, and selling it used is part of the cpu value that intel has denied its customers.
I highly recommend closing Chrome before running test. It can eat up CPU resources especially if there's a bunch of tabs open. Update: Ideally, you shouldn't have anything running to the best of your ability. Obvious things like Windows services are fine, but avoid running things like torrent downloads, Steam downloads, your browser, or anything similar. This will give you a Safe Mode-like experience to test the CPU.
i think people only kept their PC's on because powering up and powering down might be detrimental to certain components.. like mechanical harddrives. but there are fewer and fewer mechanical harddrives being used these days. people might have 1 large one just for general purpose stuff and things that dont need fast loading.
Not just that, if you live in a colder climate on and off causes stress on the motherboard and other components expansion and contraction. A pc left on is generally more reliable. It's like a car that does long distance start stop twice a day vs short distance with 10x start stop.
A desktop CPU should never degrade to the point of noticeable failure within the useful lifetime of that processor. I still have a Core2Quad Q6600 that works fine, and that CPU spent it's whole life running a 40% overclock!
I have an i5 2500k that ran 1.45v 4 something ghz for ~2 years (I wish I remembered that cpus clock speed) and a Xeon X5670 that ran 1.5v 4.7ghz also for ~2 years. Both are on a shelf in my room, fully functional still. Granted who knows if either degraded faster than normal since they were only in service a couple years, but at the very least they hadn't degraded enough to need loss clock or more voltage
@@Codyslx yeah I've been seeing those voltages one newer Intel cpus and always felt like it HAD to be way too high at full load for a stock config, but ive been on Ryzen for probably 4 years maybe even 5 now so I thought maybe I had just lost touch with Intel stuff and that they truly could take that voltage and heat 24/7 these days. Clearly not and there was actually a reason for concern xD
You don't need to OC computer components anymore they are so fast now out of the box that OC is a waste just throwing more heat at it for no real gains and in gaming its just dumb to overclock for 5 fps gains or even less specially if you have a 4090 GPU its beast at default speeds why OC it. For a basic user that just games on there computer OC anything is a waste minimal gains for way more heat and power use is not worth it imo.
Fully agree. I do the same, PLUS, I use the best cooling components that I can use. Cold electronic components last longer & are more stable than those subjected to an ongoing blast furnace of heat. Sort of a trait that I picked up from working in large mainframe computer shops long, long ago....
None of these will surely tell you if you have a CPU problem. Apps crashed constantly on my 13900K and OCCT didn't show any errors on 3 different CPU tests (1 hours each). Also, other CPU benchmark tools didn't crash. But seller confirmed CPU degradation, and Intel confirmed it. Now waiting for a new CPU to be delivered (was under warranty).
I bought an Omen 45l with i7-14700k and it was DOA. I returned that and bought an Omen 40l with i9-13900KF . After updating it worked one day and then black screened. Totally unresponsive to any key commands. Unit returned today. So scared to even try to purchase another gaming pc. Thank you for sharing how to check your cpu. I'm definitely not buying another Intel product.
As much as Intel is telling everyone to do a Bios update to fix the issue, there are millions of users that are not comfortable or do not have the slightest idea of how to do or want to take the chance of bricking the PC doing a bios update. I don't feel like that is a viable solution to the issue for 90% of the people that have a 13th or 14th gen Intel CPU. Even as somone who has built all my own systems, I won't do a bios update unless there is a way to flip back if something goes wrong. I have a Lenvo Laptop that is now a replacement due to it locking up when doing a bios update using the Lenovo updater that was installed and kept notifyingme of the update everytime I turned it on, so I did it and it locked up mid update and bricked itself.
day 1 a week ago. Date of purchase? What is the exact issue? How long have you been using the system? Is the system working before? Were there any recent changes made to the system, such as hardware or software before the issue occurred? What troubleshooting steps have you tried to isolate the issue? Are there any signs of physical damage in the unit? Did you overclock the CPU? What is the exact motherboard model used to run the system?
My 14700KF temp wreaking havoc rn I can't tell if maybe just my case doesn't have as good of an airflow or there's something wrong with my aio but I had to disable turbo to stay at extremely low temp cus its keeping my clock speed really slow too (but no effect to gaming since i get high fps on most games even with no turbo) but the 0x129 bios update essentially did nothing for me, sure maybe no crashes but a cpu thats always heavy cooling dependent i wished i had known before i got this haha
@@Lawfyyother than the obvious physical checks its probable that your specific intel defaults spec dosent undervolt automatically when posible its one of the reassons why i still use the non default presets is because they auto undervolt and get better performance for lower temps howerver obviously this can vary a lot across motherboard brands but gigabyte has been good for me during this insanely overexagerated chaos
I purchased my CPU in January, died in June and took my motherboard with it. I now have a new MSI board and CPU, it took 2 months to complete my RMA's with everything going on. A positive note the latest BETA bios from MSI with the microcode update, has STABILIZED my CPU FINALLY! Prior I had to underclock my CPU at a ratio of 6! In order for it to work properly.
Ahh I love the irony, I still have a 4790k myself, replaced it with a 13700k back in 2022 since it was starting to show its age playing newer games and trying to record at the same time. This whole ordeal was a slap on the face, I did got it replaced but it made me re think why did I even upgraded lol
I must have won the undervolting CPU lottery because I've been running my I5 13600k with CPU Lite Lode set to 5 on an MSI Pro A Z690 MB for the last year with no stability problems at all and that includes Ycruncher, Cinebench and Prime95. In CinebenchR23 it pulls 153w with all cores at max turbo and temps never go over around 65~70c
did you change anything else such as cpu core voltage? some on here recommend to set it to 1.25v with llc mode 3, or 1.3v with llc mode 5, but both may be higher temps when idle
@@KnightmareUSA No just "CPU lite lode" mode (which is MSI's "one stop easy undervolting" feature and disabling I.A CEP (which used to default to off but now defaults to on in the latest MSI bios). If you don't disable CEP it and the core voltage gets "too low" then the CPU will start clock stretching as a measure against crashing, however this is very over sensitive and tends to clock stretch well before it really needs to. You can do a little better with trial and error and adjusting various settings manually but the lite lode setting is basically a table of settings controlled by one "master" value and tweaked by MSI to get good results. With this set to 5 on my i5 13600K CPU core voltage is typically very low and never goes above 1.2v, idle typically is around 0.8v ~ 0.9v
@@atomicskull6405 I have the carbon z690 in which cinebench and prime would reach high temperatures without using one of the settings I mention. The difference is that when idle the system may have a slightly higher temp by about a degree or two, but it's very hard to tell as the room temp varies. Setting to a constant 1.3v perhaps might also annoy some who wish to save energy and prevent wear. I'll try resetting the bios settings and trying different lite lode settings but I suspect the cpu temps will still remain high. I think the high temps when stressed might also relate more to the motherboard than just it's bios, although when in normal use and not testing the temps are fairly low (in the 40s for gpu, between 40-50s for cpu)
Man i just learned like two days ago that my undervolt hasn't been running for maybe a year... no wonder my laptop was getting so hot especially during the summer lmao. Thankfully turning all the vm stuff off fixed that and my new -175mv core -70mv cache undervolt is surprisingly stable.
Updated to the new Asus BIOS with the beta microcode the other day. Was being sparing with my PC (13900K) since this all came out. Using Intel limits and I started getting crashes with Diablo 4 after playing for close to 10 minutes. The try to boot up The First Descendant and I get a “Out of VRAM” crash. Welp here we go…
I had similar issues when I upgraded to the new BIOS. Check if your DRAM voltage is sitting on where your RAM manufacturer has tested the RAM to be stable (1.35 V for example, just find your exact RAM DIMM model and see on the manufacturers page, this only applies if you are using OC RAM ofc). And check if your Vcore voltage is potentially sitting too far below your VID voltages. If the temps are fine on your system you could just - for trial and error - go and set a slight OVERvolt (offset mode) to your Vcore voltage, something like 0.01 or 0.02 V while keeping the core temps in check. And do try the OCCT variable CPU + RAM stress testing, its very handy for checking if your system is stable. And of course keep playing Diablo to see if the crashes disappear.
I have an i913900K and ASUS motherboard. I'm not very computer savvy... The new non-beta BIOS update (version 1663 including the 0x129) from ASUS says "for non-K processors"... and I have a K processor. Do I need to update to the version 1663 BIOS? Last month I updated to the older 1661 version which states: "Updated with microcode 0x125 to ensure eTVB operates within Intel specifications." Never had any crashes or trouble. I'm a little confused about the "for non-K processors".
@@turtle2720 I think it just adds the microcode for the non k processors as well, it used to be only for the k processors. so i do´nt think it would change anything for you but it probably wont hurt to install it anyways.
@jeremyashcraft2053 I was not funny enough. Everything was stable and fine about a month ago which was my last big play session. Maybe something in the new microcode and BIOS isn’t playing well together. I’m going to downgrade my system to a 12900K for now and wait for this to all blow over.
I get a little frustrated when I try to watch these videos and I can't understand them, but hey, here we go, trying to learn a little bit every day.
The same thing happened to me, but after some time and studying, I am understanding all these things. Now I am in the process of understanding the performance differences between different versions of Windows.
After searching for my Windows 11 on BNH Software, I understood that on Intel processors it is a little better in terms of performance.
Did no one think of just asking the CPU how its feeling?
Lived experience is important.
@nezbrun872 they're all missing the comfort blanket of dust and cat fur from days of build up in a cat ladys home
This was brought up during the DNC.
My cpu identifies as a cat
I don't have the hours to hear how my CPU is feeling, I once did that to a blue haired college at work. NEVER again
So Intel 13th and 14th gen dying and AMD 9000 series CPUs flopped.
What a time to be a PC enthusiast.
So hence both sucks 😂
There is a windows update for the 9000 series. Its confirmed by AMD. They ran their tests on a specific Windows 11 version which can be downloaded, they posted an article on their site
ill stick with 7800x3d for next rig
@@Almin88 That's a universal CPU 'upgrade' though, it helps out Intel CPUs too in a similar way, nothing to do with AMD
And just before that Nvidia and GPUs in generally weren’t sold to gamers. Rather bots to farm fake money that didn’t go anywhere but into a few folks pockets and well of course the gpu manufacturers 😂
2008 intel silicon lottery to see if you can get your i7 up to 4, 5, or 6 ghz
2024 intel silicon lottery to see if you don't have to down spec your cpu to keep it alive
that's what happens when intel just factory overclocks your cpu instead of actually innovating
@@DeeSnow97 Alder Lake and Raptor lake, the biggest architectural changes to Intel CPUs since 2006, not an innovation, eh?
If it starts showing symptoms of degradation, underclocking it will not 'keep it alive'. RMA it.
@@Amaeyth if you put it that way it sounds better than the only meaningful architectural jump since 2015. that's the problem here
intel's 6th, 7th, 8th, 9th, and 10th "generations" are the same architecture on slightly better process nodes, with core and clock creeps. 11th gen was a lazy attempt at a slightly wider core from the previous years, backported to 14nm. 12th gen was intel's only jump in the ryzen era, and then 13th gen was a minor cache bump, and we're back to stagnation with the 14th gen.
that's quite a terrible track record for the past eight years as far as innovation goes. six of these eight "generations" are just marketing, nothing more.
Blaming mobo manufacturers is wild.
K, KF, and KS model CPUs are "unlocked" yet apparently any kind of overclock whatsoever just fries the CPU in a matter of months, while the current Intel stock values will likely fry it in under a year.
We are going back in time when installing a component is no longer as simple as "plug n' play" like back when installing a new HDD required you to configure the switches on the HDD and go into the bios to tell it what switch it's set to, oh and also make sure is in the right position on the IDE cable as well.
Except with my 14900KF, I need to update and reconfigure my BIOS each month, but hey, at least I don't require an Intel account to run my Intel CPU.
Yet...
We should at least get a 25% cashback considering the unlocked models are not safe to overclock, and as such, we paid a premium for a feature the CPU wasn't built to withstand, like a Wish product that only lasts a few months.
Step 1: Download the free Black Myth Wukong benchmark tool
Step 2: compile shaders
And?
@@rotormannwait for crash, also works with first descendant 😊
@@rotormann BSOD or not BSOD
@@Pow3rus there is another: out of video memory 😉
@@marcinkarpiuk7797 Ok. No crash for me. My system is very stable.
Fobbing away from the fact that Intel should be FORCED to swap out all CPUs from the affected production runs last year etc.
Hopefully US top notch lawyer's can bring a class action against Intel ( with EU smacking Intel a bit ) to offer all affected customers (of 13th - 14th Gen I7-19 Chips) - Equivalent of their New CPUs (15th Gen 1851 Socket Intel Arrow Lake ) and $200 - $500 for an equivalent Motherboard to house their new chip CPu. All this f^^&^&^& up is harming the Intel Brand and p^^%%^^% lot of us off; with view that our new build will unlikely be Intel but their main rival AMD; many say it's like the CPU equivalent to B loeing 737 max f&^^&^& up.
@@dafyddthomas7299 they probably won't, for two reasons: intel also has top notch american lawyers (arguably even more so because the us operates on the doctrine of the rich shall eat the weak) and because splitting intel apart and declaring bankruptcy on the cpu design business would probably be cheaper than providing all that stuff.
Intel garbage.
They should do a new run and fix them in the process.
A new CPU (15th gen ?) means a new motherboard
@@DeeSnow97 Yeah, in hindsight them announcing the foundry split is super sus. Perhaps they knew it way before they let on.
Timestamps would be nice :)
The days where you watch the video from beginning to end and comment later are gone 😂
@Ladioz I wasn't planning on watching the video since I don't have a crashing Intel CPU, but for the people having it, it would be nice.
0:01 video start
@@FantasticHydra I look forward to the timestamps on your videos - when u put the effort in
@Custmzir, I appreciate it. 😅
It would have been good seeing a degraded system be tested
I've heard claims of anywhere from 50% to 99.9% failure rates with not a single example being shown
My 13900k is degraded. Can't pass 28000 on cinebench r23 with my Asus board on the new microcode. Even with extra power I can't pass 30k. Waiting on Intel with an rma
I would love to send my -cannot surpass 190W CPU- to Jay, but I am absolutely dependent on this system every day.
He mentions that he has a degraded one that’s shipping to him. Agree this will be a great video and looking forward to seeing it
@@AZGhost623 for sure. I can get 32-33k on the mobile variant. Hope they deal with it quickly.
If your chip has degraded it will require more voltage for the same clocks. Basically the quality of the silicon will diminish and it will become less power efficient.
Other damage from degrading could include instability so bad that voltage wont fix it, this is the most severe of damage and requires RMA.
When you reduce power on a degraded CPU to curb the issue, it may only cause more crashing. This is due to the CPU now requiring more power than when you bought it.
You will in most cases have to dial everything down but then up the vcore until it stops erroring / crashing.
Disable multicore enhancements. Use the balanced power plan, enable C-states. Stick to a MAX of 1.35v vcore and monitor this with OCCT, HWINFO64, etc.
Lock power limits to 253w for PL1 and PL2 and lock Amps to 307A or whatever you want below that.
You can set adaptive voltage offsets with a minus symbol to further turn down the power your CPU draws, other than simply reducing vcore on its own.
Another solution would be to lock your cores so that you disable the high power draw "6ghz boost on two cores" BS.
Wouldn't that mean raising the floor? Capping the ceiling at what is still ridiculous (1.55V) isn't necessarily reducing voltage.
@@douglasmurphy3266 What the hell are you talking about?! I said 1.35v not 1.55v.
@@nontoxic9960 Yes but Intel are saying 1.55v is "good", as its baked into their profiles.
@@Intel13thGenBurnsand you believe that? FR?! Do it then and see how long your cpu will last.
That's still insane how much you have to do just for your CPU to not kill or injure itself. Most of us like simple things like a quick BIOS update.
"There's a timer on your CPU and you should turn it off more!"
Sorry, I'm used to running CPUs so long that they're hopelessly outdated before they run into degradation issues. I was still running a 5000+BE in 2016, and it still did everything it once did. Yes, it's electronics, yes, electronics have lifetimes, and yes, the 13th/14th gens are trying to kill themselves at what should be but isn't an idle state, but it's still quite shocking (pun intended) that we're now seeing stuff go to hell quickly enough that the replacements don't exist yet.
Same thing. Many old processors, even with OC lived for so long they simply became outdated, suitable for people who cant afford anything better. CPU dying that fast isnt norm, just as it isnt norm for GPU`s to fry themselves without doing something horribly wrong (like using molex adapter to 6, or, god forbid, 8 pin connector).
I'm a prime example of that as proof.
I've been too damn lazy to upgarde my machine. My machine is a Core i7 965 first gen Nehalem. I've left this thing on all day for over 16 years. Granted, I never took major advantage of the unlocked multiplier or its overclock capabilities, but this thing has been running in a place that has climate very similar to Florida's (80-90F at worst, but high humidity) with moderate air cooling.
And it still runs, it still doesn't BSOD, it still has no problems running the games it can actually run.
Like you said it's depressing AF that my i7 965 from 16 years ago running all day in warm and not-so-ideal conditions is lasting longer than these 13/14th gen Intels.
These things are literal ticking time bombs and Intel should be ashamed of their screw up.
its not only degradation. they had production issues. so most of the cpu affected are affected by this.
above that probably a managment problem forcing the stuff to crank up the voltages and power consumption to insane levels to be able to keep up with amd in synthetic benchmarks.
literally the first thing i did with my 13700 when it was new, is that i lowered the power limit in bios to 160 watts, and lowered the adaptive core voltage. my processor still beats prime 95 no problems.
I'm still using my Intel i7-2600 CPU @ 3.4GHz I purchased in 2013, which I still used today. The PC has been left on 24/7 all those years. The only upgrades have been an SSD and an extra 8GB of ram. I'm looking at building a 4K editing machine with an i7-14700K and a 4080, but the constant talk of stability issues is preventing me making the purchase. Maybe I need to get over the fact that it won't last another 11 years.
If degradation of CPUs was so bad people couldn't still be run 486 retro Pc some 30 years later.
I have had my 13900K running on air since the day I got it (I know, I know). It had these same crashing issues since day one. So, from the very beginning my goal was to get acceptable temps. So I immediately did some BIOS tweaks, undervolted, lower maximum clocks, LLC adjustments, etc etc. With the tweaks I ended up making, my 13900K maxes out at 80C under full load on air. It also eliminated the crashing issues. The new BIOS runs perfectly for me without any tweaks, aside from temps, so I applied similar changes I did before to lower my temps and I'm pretty happy with the results. In a way, I feel like I, "saved," my CPU just by having it on air (against recommendations) because it required me to undervolt/underclock right away.
what is your cooler? maybe you need to replace cooler, if you have bad cooler, it may stop. i have 13700, and 14900. I replaced with highest cooler out there.
I underclocked mine to 3Ghz all-core and it reaches 35c max.
@@mikeree8543 It was a choice to run it on air, I knew I'd have to undervolt a little to maintain manageable temps
@@dxncx4555 You're missing out on 2.8ghz of performance though. Why not keep raising those clock speeds until your temps sit just under 80c while under stress
@@durrik it was a joke :D Running it Stock with some tweaks for better Temps, also installed a contact frame
It is awful that Intel Customers are put into the position that they need to do this stuff.. !
Maybe just LIE like AMD ?
@@tilapiadave3234 Not quite the same. AMD's lies are about performance. Getting your AMD CPU and finding out it is slower than advertised is not the same as getting your Intel CPU and finding out it will be dead in three months......and that they might not replace it for you.
@@tilapiadave3234lie about what again? Is it the recent performance issues with the 9000 series? At least the problem was identified by reviewers before the launch, unlike Intel's issue, which took over a year to show signs of degradation after they sold a lot of their hardware to consumers, enterprises, and data centers. I would still prefer AMD's lies on performance charts over Intel's unfixable and irreversible damage, which they tried to sugarcoat you with a BIOS update. Intel knows that the issues lie within their hardware and not the BIOS. Yeah... Yikes.
@@mrpekko98 intel expand their warranty and u can RMA your cpu if its damaged so no need to cry AMD fanboy
meh, not like it's never happened before with other chip manufacturers. Stuff happens. As long as they don't do it again, whatever
Ugh. Got my 13700k in November of 2022 a few weeks after it came out. I was SO hyped to have a next gen CPU. My PC started having bluescreens about a year ago and they keep getting more and more frequent. BIOS updates, disabling XMP, reinstalling windows, nothing has worked. I am in the process of RMAing it now. Such a shame.
It’s crazy how you talked about Falcon NW right away because I had a 45 minute phone call with them where they walked me through the bios update and profile preferences. Thanks for promoting them on your channel, they’re awesome!
I bought my 13700k around launch back in 2022 and yeah the degradation started around 7 to 8 months in, started crashing while compiling shaders and such, specially on emulators.
I RMA'd mine and to be fair the shipping and replacement arrived rather quickly, HOWEVER, be aware that in the email they send you with information about your RMA and about your replacement, it will state that the model they'll send you will vary depending on what's available, I got in return a 13700KF instead, I personally didn't mind much since I'll never use the integrated graphics and have spare GPUs in case of emergency, but keep that in mind if you do RMA your CPU.
@@1newme425 See, that's what I thought as well, and I agree, sadly I needed the CPU back as soon as possible as I had work needed to be done, don't have any other working PCs atm, loosing that part of the CPU was ok just to resume my work.
That's why I posted this so people are aware and call intel out if they notice a smilar RMA.
Many people who RMA’ed their i9-13900k’s and got an upgrade to i9-14900k’s.
Meanwhile they technically downgrade you. They gave you a product lesser than the one you had originally. That’s absurd.
But you're looking at VID, not VCORE. VID is what is requested to the CPU, not actually what is reaching it...
Don’t know why people still listen to jays hardware advice. He’s a cool guy, but he just has no idea what he’s talking about when it comes to this sort of thing.
@@gildfps Agreed. Grew up watching him, but I feel like he should educate himself a bit more before posting these types of videos.
Yea, he needs to be looking at VROUT. At least if everything is set correctly though, VID should be close to VROUT.
Yeah this tripped me up with undervolting, until I learned that you want to look at VCORE, not VID.
I know right that was bothering me the whole video. Then again maybe he's smarter than we think and he's aligned his loadlines just right so they are the same. But probably not....
EDIT: People seem to not be reading the last sentence of my post. I don't think this is the case, which is what "but probably not" is supposed to mean.
Step 1: low gpu usage beyond normal?
Step 2: random lock ups even at idle
Step 3: USB flash drives randomly go offline or not recognized
Step 4: Any bluescreen?
Step 5: heavy crashing Unreal 4 or 5 Engines
Step 6: Out of Video Memory error when starting Red D Redemption 2
Step 7: Random restarts without any overheating or lock up
This is what I experienced on both of my failed 13900k cpus. That was also undervolted, watt limited, and cores locked
Never drew more than 1.35volts in any scenario. CPUs began showing 1 or 2 symptoms above and eventually all of them. Was happening at 6 months use marker on both cpus.
It is a shame as 13900k is amazing for workstation and gaming, but I am already tech level gut that is dealing with isolating issues with cpu. It happens again at 6 months marker. This repeatability is an identifier of identical issues within the cpus' designs. Same failure points. I never encountered same identical issues in 25 years of building pcs, down to amount of months
I'm seeing numbers 4,5,6 with my 7950x cpu. Which is crazy because my i7 4790k is far more stable than this...
Well shit, I have both step 2 and 3... am I f'd haha?
MY 13900K died, I think it degraded enough and intel immediately approved RMA lol
How long did the rma take? I put a ticket in on Sunday and it's been radio silence since
Lucky, they denied mine
Should of got a KF version....
@ZenRyoku what did that matter? One has integrated video, one doesn't...
@@SwingArmCity I was being feciecous...fukin around...
I actually have a an i7-133900-KF maybe misplaced a 3 or 9 (or whatever) basically the processor KF vs just the K version is supposedly completely unlocked with no restrictions on it... it's been a while since I bought my latest pc...so forgive me if im not 100% accurate...
But these damn octocore 12th gen processors sometimes acheive almost 5k htz range...which is fucking insane...i started underclocking for the sake of how damn hot it got in my room..lol
I even removed the glass plate that is the removable side of the case in case you got to get inside for anything and replaced it with some very cool perforated screen specifically to protect against dust, but let's the entire pc breathe and disperse enormous amounts of heat to which I added a mini tower fan about 1 ft tall that helps ventilate the pc as well as the area around it... Much better than liquid cooled stuff in my opinion...
No matter how hard i run my pc (also has a RTX 3060 ti (the bad ass kind ppl use for mining bitcoin), plus the 1000 watt power supply....
...the processors or GPU never get above 80 degrees centigrade....which most of the time operates in the 60-70 degrees range on tha reg...
Which accomplishes (1) The sheer heat the entire rig typically gives off.. (2) allows the PC to run as hard as i want it to without overheating...(3) helps stave off degregation for much longer than running sealed in it's fusion level hot generator sealed metal box compartment....
Anything electronics that run with capability and intention for performance will ALWAYS break down over time....it's physics...💯 THERMODYNAMICS ARE VIRTUALLY IMPOSSIBLE TO CIRCUMVENT.... EVERYTHING WILL DEGREGATE or deteriorate over time....
But the only way to extend the life of your performance based electronics (specifically PCs) you just have to get creative or specific with the hardware you decide on to get the most viable life out of it
Had my 13900k for almost 2 years no blue screens, freezing or crashing. Usually upgrade every 2 years anyway so I'm not going to lose any sleep over it.
Like you I am on a 2 year cycle as well, but on a 1 generation behind schedule. Meaning I dont go with the newest generation when I upgrade I go one generation behind, that way I miss all the growing pains of that newest generation.
Same. I've had no stability issues whatsoever. I get more problems from NVIDIA 4900 drivers. 😉
Don't Jinx it!
2 years also did my 0x129 bios update…. My cinebench score is 34000ish on multi core for 10 minutes… not a good score and idk why
@@KingDooksare you thermal throttling? My 13900k actually gained about 1500 points after undervolting. R23 score was around 34k, now it's passed 36k. It somehow runs at higher clocks with less voltage. And I'm also not using an unlimited power limit which is how most people get really high scores.
I am really surprised by my midrange Asrock board. First vendor that did not pump insane voltages on auto settings. Asrock is by far the best board manufacturer. Okay maybe I got a 1% CPU with an insanely low VIP table, but Asrock truly rocks. I am an Asrock fanboy now.
Just wish they would update their bios GUI, it looks ancient. But still love my steel legend.
Electronic components initally powering on is hardest most stressful thing it can do. This is why when a light bulb bursts, its typically as you flick the light switch (though now-a-days LED bulbs are a bit different but same principle applies). So personally I wouldn't adhere to that advice at the end about turning off your PC all the time that you're away from it. You'll potentially cause your CPU to degrade faster, along with every other component in your PC, depending on how often you actually use your computer. As someone who uses their computer every day, i almost never shut off my PC. I have an i7-950 from yesteryear that still functions today, and was something I pretty much ran 24/7 for nearly a decade.
This is simply a myth that was debunked years ago.
I always turn my pc at least once a Month and my Last pc wish was a laptop lasted for 10 years
I have an i9-13900. I never had any issues with stability or crashing, but I upgraded my bios to the latest to ensure I had the new Intel microcode. For the first time ever, Overwatch crashed on me yesterday. I'm concerned the new BIOS is actually causing me issues, but I only have the single crash so far.
Worry reduces lifespan.
Live longer with AMD
that means you processor has degraded, you will need to add a low off set and lock the cores, the reason why the new microcode is crashing it is because its too late and the undervolting is triggering the crash.
I’ve had the i7-13700F for about 7 months now and have zero issues whatsoever with it and been hesitant to update my bio for the new version for that reason as I’m not sure if I should really do it if I’m not having issues
@@bardbacx I suspect the BIOS was rushed out and has bugs. If the CPU was degraded I would have expected stability issues prior to the "fix"". Thanks for the input though.
My guy, don't be paronoid. Remember programs would crash randomly even before this issue. A single crash doesn't mean your cpu is on its death bed.
Me talking to chip: "Show me on this doll where Intel hurt you?"
Built a PC less than a week ago. MSI MAG B760M WiFi II motherboard, gen 12 12700k, 64 GB Corsair Vengeance RAM, MSI RTX 4060, Corsair SF850L PSU, MSI 240 AIO, MSI 500 M2, 2TB 2.5 platter backup drive crammed in to a Jonsbo Z20 case. Yep, I had RAM OC errors before BIOS update. Not currently. Runs smoothly, for now. Love it! I'm not an overclocker or heavy gamer. Videos like yours kept me motivated.
My blue screens were caused by my boot drive getting too hot.
It took a long time for me to figure that out. It started the moment I swapped to the 13600K from the 5600g.
It’s definitely the difference in power draw, so the 13600K ran hotter.
I fixed my problem after swapping to a new case.
I just switched my i9 14900k for an i9 12900ks and I'm loving it. I'll probably switch to AMD next build but the 12900ks runs like an absolute dream and is a beast. Such a shame that Intel have blundered this so damn hard.
I happened to switch right before all this went down. Got lucky I suppose. If you do, dont be like me and forget you have to flash amd bios to get it running. Man I spend way too long troubleshooting nothing.
@@Stony1313 There is a list of steps for EVERY CPU and new motherboard. (And of course, one for EVERY used CPU/Motherboard combination.) If you stick to a plan every time you get a "new" system, you never waste time. :) I know it takes (upfront) time, but it saves hours troubleshooting.
Yeah I switched to team Red for the first time in my life, right before this crisis hit! For once in my life, something worked in my benefit! 😁
In comparison, I once bought a v8 Chevy. The next day America announced sanctions with Russia due to Ukranian war. Making gasoline skyrocket in price! 🤦♂
I have an I9 13900k I've been using since May of 2023. Not sure if it is degraded or not but I really don't want to do all those tests to find out. If it craps out one day I'll just replace it, no big deal.
My thought exactly use it tell it stops working
If you get any out of video memory errors, it's your cpu. Had my 13900k for about the same time and had these errors start about a month ago. Now it's fully bricked.
Good luck!
Dear Mr. JayzTwoCents, I just want to say thank you for all you do. Your videos have been very helpful when I'm trying to figure out something with my PC.
I'm still on the fence about upgrading the BIOS for my 13600K on a Gigabyte motherboard.
It's super stable, runs under 1.25v max with temps at ~72C while gaming. I've seen only 2 games that push the power over 100W (excluding shaders compilations).
Better wait for a non beta bios anyway, it really does nothing for the 13600k except introduce a new intel default that kinda sucks on efficiency.
i have same experience here. I undervolted almost right out the box. Still hits 5100mhz boost 65C @1.25v iRacing in VR is just using 80W
13600k here too. Not seeing any issues, not going to update to new BIOS because of performance issues. If it dies I'll go to Intel I guess. I got mine in 2022 though. It was speculated that the original first batch don't have this issue. Has anyone else heard that?
If you do get it, make sure you set the power limits and ia voltage limit anyway, unless you -really- trust gigabyte and intel.
im on Bios version 2305 with ASUS z690-F and I have spent some hours some days and cant see any voltage problems or anything, I have HWmonitor and tested with Cinebch and voltage dont go over 1.251. I do XMP 2 and only lower offset voltage, never touched LLc or anything else, I run my computer stable, never crashes im on 13600KF
Jay this video answers every question I've had for the last 3 months. You are hands down the best RUclips channel around. Thank you so much
Intel RMA Rates just went up 90 percent after this video 😎😎🤭
There is a long line ahead of them. i've been waiting 32 days already with no refund.
People are starting to realize that they can't "fix" the issues with micro coding updates.
@@jabronilifestyle lmfao
@@bjarnis Can't fix the damage that's already done, but has the actual issue that caused all this been fixed? Seemingly yes
And most of them will get there cpu sent straight back and charged a disgnostic fee. Be damn sure you need rma before you kick and scream over the phone.
FYI, I have been having constant crashes when running Diablo 4. I tried running Starfield and it Crashes.I tested my new NZXT-PC with a i9-14900KF and both Cinebenchs used in your video and they crash and OCCT wouldn't even start, and when I ran Intel's Diagnostic (latest version from Intel), the Burn Test, Unigen and NVIDIA install had no issues. I swapped out my Gigabyte RTX4090 for my ASUS Strix RTX2080Ti and no crashes. Clearly I am having video issues. Thanks for the great video in helping me identify my GPU issue. Stay strong and thanks for the great instructional video.
the only intel I have is at work. It's a dell. It's in the affected list. I'm hoping that dell being crap means that it was never hit hard enough to worry about. It took forever to upgrade from my 4th gen CPU. They really want laptops here. I HATE laptops.
It's long been understood that 70 degrees C is is safe sustained operating temperature to prevent circuit degradation. Never had a CPU fail for over 10 years use 24/7 adhering to that. Aim to keep the top temps below 80 degrees C under load.
1st thing I did when I got my CPU was undervoltage it to 1.2V. It has been super stable and the temps are great. Stock voltage in these generation is crazy, temps are crazy, everything is kind of insane. I guess is all for the bench wars.
I just under volted to 1.250 and it went back up to 1.450 when Windows loaded. I think I’m just going to do an off set in the bios to get it below 1.350.
XTU doesn’t allow me to undervolt since this new bios update.
14900k.
The problem is, that Intel should set the stock voltage at healthy levels (and clockspeed at stable level accordingly). And let people themselves take responsibility if they want to OC / OV the CPUs.
Intel didn't really needed to be Faster than AMD. (They already were the most popular) They got that succes by being the most stable and easy to use platform.
They managed to survive and even grow with the dreadful "slow but stable" Pentium 4 against the faster Athlon for years.
Very good video. I built 3 custom systems spread over the last 18-20 months for a local business. A 12th gen and 2 13th gens (all i9's, high end builds). The last 13th gen i9 failed after 2 months of use (at a business used 5 days per week, min 8hrs per day). This happened before it became widely known about the issues with Intel. I RMA'd the motherboard and CPU just to be safe, and I didn't know which component caused the failure (to simply turn on). BTW, the replacements worked. I have told the business to check their MB settings and update the BIOS on those boards.on several occasions. I don't think their IT person has done either. I built myself an i5 13th gen and have set the motherboard to the correct default setting, for a while now. All thanks to your videos and information.
Intel offered me a refund as they don’t have any replacements in stock for the next 4-5 months per RMA email.
Take it and put 12th gen bro
I was offered the same they then came back to me to say they have some new CPU's available. Im still requesting the refund (Told its going through the approval process in upper management) I will be going with the 12900K.
🤣
@@mooshi-x7927 is 12th gen supported on x790 mobo?
@@richardkeller5703 12900K is virtually identical to 13700K. I bet you, that CPU is affected as well.
… Brilliant … thanks for going through these tests with us … as ever you are responding to the needs of your community👏
At no point did you check the actual realtime voltage until you glazed over it in OCCT. Pretty sure MSI boards use die sense so you should be able to see the voltage the CPU is actually receiving. The VID and the actual Vcore are not one and the same. Also you neglected to mention that current draw is the main factor in electromigration, and not voltage. Also the instability in shader compilation people have seeing is very much an all core, current intensive workload. A lot of misconceptions here that aren't quite right.
bro just does not know what he's talking about.
1. This is for all boards not just MSI
2. Voltage/Resistance = Current (it's called Ohm's Law)
Thanks for explaining this with sentences. Others were giving 1 liners on the subject, short sentences and not saying enough to be worth my time to glance over their comments! Everyone upvote useful comments like this one please!
@@yonghominale8884 The MSI comment is in regards to die sense measurement using differential sensing pins being used and not meaning to say it's only applicable to MSI boards.
Yes, however simply saying "it's ohm's law" doesn't help this video as it's simply saying voltage is the primary cause. We're not even mentioning the impact CPU load load has, or the fact that different loads cause different current patterns.
Had same issues you spoke about for my 14900k during first 3 months of 14th gen release.
z790 hero / z790 strix e
Artic liquid freezer II 360
Used two different motherboards and two different 14900k chips. Both scenarios produced the same results. Cinebench and games would crash upon starting and during cpu high spikes.
Everything was tested stock settings out the box.
Returned both chips and went with a 13700k. Everything has been running stable since.
Thanks JayZ because before this whole incident happened you post a video about motherboard manufacturer made a mistake on having a higher power profile than intel intended also provided us a guide to undervolt to mitigate the instability issue that's why I able to dodge the bullet on damaging my CPU.
He was wrong to blame the motherboard manufacturers for intel's faulty microcode.
@@kriszhao80 Ofc the fault is on both of them if motherboard manufacturers didn't put the unlimited wattage the volts will not go higher. And it will not trigger the problem.
@pacspacs8693 Intel literally sells unlocked processors as a selling point saying that they are designed to be overclocked and has been using the overlclocked benchmarks for years to try to be competitive, you can't promote the behavior then when you make a faulty product blame it on the motherboards who you have been encouraging for years.
@@kriszhao80 Exactly, and yet Jay goes out of his way every time he talks about this issue to throw shade at the MB partners. None of this is their fault. Intel just looked to shift the blame before finally admitting that it was their borked microcode that was 100% at fault for these problems.
@markh4750 yeah I remember him doing a video about a month ago and he blamed the motherboard partners instead of Intel.
Hi Jay, great work as always. As an RF, Analog, Mixed-Signal CMOS integrated circuit (IC) design and layout Engineer, I can tell you that, in general, the Silicon Chips, including cell-phone, desktop and laptop communication chips and processors are all designed for a 10 year lifespan with the highest currents and the resulting electromigration. Some last longer, some last shorter, depending on the process nodes. The computer I'm using to write this message is a 14 year old HP Pavilion with Intel's original i7 860 processor with NVIDIA GeForce GT 730 (that was installed later to allow for 4K display - since I'm using my 55" UHD TV as my monitor). I do turn off this computer between 8 to 16 hours a day.
The first and only gaming computer that I built with your help and advice (thank you Jay - and Linus Tech Tips) uses i9 13900K CPU (liquid cooled) and NVIDIA 4090 MSI SUPRIM liquid cooled GPU. We'll have to see how long that will last.
Built in Feb 2022, my 13900k/4090 has not degraded. But only because I believe I took the precautions at the time an extra step than even content creators in that era had recommended. They called for PL1=PL2 limits of 253W. I set mine to 125W instead. And ran the 4090 with power limits at 70% in Afterburner. With enhanced turbo off and a -0.08V undervolt. My friends were curious as to why I chose to run everything so low. At the time, it was simply because the heat of sitting next to my pc was uncomfortable. So in the 80C and even higher 70C range, that was already too much for me. My settings allowed me to use my computer at 50C and 60C instead. At the cost of some performance.
Kinda sad though... investing so heavy in a system to then downclock it, but hey happy that your CPU is doing fine! :) My 13700k is not experiencing any issues either
buy a fast cpu, slows it down because of temp issues what kind of logic is this.
@@squallsquall9669 Some don't care about getting every little bit of speed (but close to it), some just want stability and enough speed close to what was advertised to play all the games and do all the things they need and just have it work. Also some when they bought the processor did not know they would have to do all this undervolting, changing of settings etc.. that came to light. Because all this did not come out till later.
I also have my 14700KF undervolted at -0.08v. Set PL1 135 and PL2 150 based on observation during gameplay and loading screens (time to compile shaders, etc.) The games I play typically use about 175w to either compile shaders or load from area to area, so I capped it at 150w peak for 56s and don't even notice a difference (probably milliseconds in loading). Same during gameplay, which peaks around 130w, so I capped it at 135w so it doesn't hinder performance while the undervolt keeps things cool. I generally don't see anything over 60C in the most intense scenarios (generally low 50C) during gaming while fans are set around 50%. Idles at around 30C. Volts usually peak around 1.35v at 5.5x ghz during intense gaming. If I ever need more temporary power, I'd just up the PL2 back to 253w worst case, but I'm in a good place atm.
@squallsquall9669 believe it or not, undervolting and limiting the watts only really affects benchmarking and has little to no negative drawbacks during gaming (not sure about rendering etc.). In real world gaming scenarios, dropping temps down 20C while maintaining near spot-on performance from stock settings is great. I notice absolutely no difference during gameplay and load times. It could be because I game at 4k so everything is more gpu reliant, but hey, it works perfectly as intended.
If love to see a more detailed vid with OCCT, either OC’ing, stability testing, undercoating, memory OC’ing or something to explore the tool more, there’s of options. Love the info as always Jay!
I personally would happily pay for OCCT except they want to make it a subscription. Being a subscription offends me, because it's designed to hope you forget to cancel it. And this is just for small businesses/home users. Not to mention, it does *not* include all of their enterprise commercial license agreement features. I asked the owner about that because as a paid home user I wanted to use those features, and they are not going to bring them for anyone other than enterprise.
2:17 I like the unedited reexplaination
Lot of people complain about MSI's quality, but we have this happy coincidence where MSI using lower power limits (possibly to compensate for lower quality parts) has spared their 4090s from melting connectors and now possibly spared Intel CPUs from early degradation.
The connectors have nothing to do with this.
@@johnt.848 I didn't say it did. I was commenting on how two completely different issues were _coincidentally avoided_ by having lower power limits and the reason MSI has lower power limits on both their GPUs and mobos is probably because they use cheap parts.
It's an amusing irony the brand with a deservedly mid reputation is doing better than everyone else because of this.
MSI's 4090s are also burning.
The issue is due to the connector not having a tight enough fit in which a slight gap from the cable being bent too much causes arcing, which produces enough heat to melt the connector.
@@Skelath Unless there's been new information out, from what I recall in the North Ridge and Jay's videos is that MSI is underrepresented in the available data regarding melted connectors. The only way that's possible is if a) nobody's buying MSI (unlikely, they would've disappeared before EVGA if that were the case), b) they have such an amazing RMA that nobody's sending them into third party repair shops (also unlikely, because MSI's RMA process is not amazing), or c) something about their cards coincidentally avoids the design flaw of the new connector.
@@blarghmcblarghson1903 you can literally search for "msi trio 4090 melted connector" and find several images.
Lock all the cores first at the real speed of the cpu, for example the 13900ks is 56 (5.6 ghz). Then after, set voltage to 1.345v and a negative offset starting at -75, and work you way up if you need more of an offset, you want to keep the "vid max voltage" below 1.35v as anything above is too much.
Also, power limit the cpu to 253 watts max will also help and make sure if you have Windows 11 that you turn off "core isolation" as if you don't it overrides any voltage settings, and offsets you set.
This is what has worked for me with my now 2nd 13900ks and been going good so far. Hoped this helped!
At 9:15 Jay's CPU Degraded for one second 😂😂😂 did you see it ??? lmao
Yikes... Well spotted! This definitely needs an explanation. He only has HWINFO open too...
his screen went black for a second, happens to my computer sometimes. what does it mean?
@@rleekc Ulike the original comment here, it has NOTHING to do with your CPU. This is simply the DisplayPort doing what it is supposed to be doing. DisplayPort essentually conducts syncing operations with the monitor, it's called link training. This link training can happen at any time and it is done so in order to maintain functionality of the DisplayPort link; what it is doing is reducing the link speed or increasing the link speed, which is based upon the best link speed while removing errors. Again, this is the result of a lesser quality cable if it is a regular thing (once a day perhaps). However, this is nothing to worry about at all, you can either ignore it or get a better quality DisplayPort cable.
Same, i assumed it's a video setting or something. @@rleekc
I will largely get blank screens when trying VR, and toggling between apps.some windows it wont like.
Asus @@rleekc
Great video Jay. I have a 14900K on a Gigabyte Aorus Z790 Pro X board with a 4080 and 64GB of DDR5 6000mHz Ram. I was getting the “out of video memory” errors on high shader games like Hogwarts Legacy and Horizon Forbidden West and I do video editing on Adobe Premier Pro and I could not export a video no matter how big or small my CPU would thermal throttle and crash to desktop every time I tried to export a video. Since the BIOS update with Intel Extreme Default Profile and the new Microcode, I can load and play Hogwarts and Horizon completely fine now, no more “out of video memory” errors or crashing and I can go into Adobe Premiere Pro and export all videos now, my system is so quiet and exports them successfully. The crashes I was experiencing was happening EVERY time, but since the BIOS updates, it doesn’t do it anymore so I would so based on my tests my CPU is stable and has not “advanced” degradation, these are the tests I did and I believe my system to be stable now based on my experiences.
Every CPU is degrading, and will fail one day, as the earth will stop spinning one day, and the sun will die. With 13th and 14th gen we are talking about EXTREME degradation. I have a colleague (asus z790 strix MB) whose 13th and 14th gen CPUs died after 3-4 months of normal usage. And I have i7@920 that I purchased in 2008 and gifted to a friend years agoand he is still using it. So let's not try to intentionally or not diminish how BAD the situation is.
And yes, I was an intel fan currently owning 13900k with MSI z690 unify. No problems so far for 1+ years, but after I installed the latest bios that was released today, cinebench R23 dropped from 40800 to 36500. All settings always defaults, no OC what so ever, not even XPM, because I'm always counting on the reliability instead of performance. And was willing to pay for a new motherboard every 2 cycles because "Intel is the best".
No more Intel for me.
"Pimpin' been around the since world started turning. And it's gonna keep right on turning right along with it. Until this little planet rotates off it's axis as a result of it's core overheating and explodes into cosmic dust."
- Black Dynamite
Companies make mistakes, simple as that.
@@montreauxsfor 2 generations in a row?
It is strange though, I have a 13700K and with Cinebench24 I have exact similar results before and after the 0X129 update.
I bought the same cpu around that same time, mine went down to 34500 after the update and I rarely even turned on my pc.
I needed this video! Should take a load of stress off my shoulders after the scare that my GPU pulled on me (no image output) after the 0x129 BIOS update!
Thank god for the win+ctrl+shift+B shortcut to reset my graphics drivers without the ability to see my screen
There should be a independent regulatory enquiry about this issue.
inquiry
Jays is only tech channel I watch even when it not relevant to me at time. there are other good tech channels but only one Jay
my 4790k ran like a champ for 10 years now? i dont remember shuting it down more than once or twice a year.
never thought of "shut down your computer" when you dont use it.
now with my 14900k i might start doing that lol
I never once thought "I'm gonna be away from my pc for more than half an hour, and have no need to leave it on, so better leave it running for absolutely no reason".
It used to be important but not leaving your pc on hasn't mattered for a good while. Ive left mine on for literally weeks at a time before
I love that you are showing all the programs I use everyday on User's PC's.
got a new 13700k , updated bios and changed lite load and put voltage offset and now my voltage is at 1.250 while at load and its super stable and cool. Was a big upgrade from my 10700k in gaming.
nice man I’m getting one soon is there a tutorial anywhere I can find to update bios and settings I may have to change?
@@therealsentinal4519 Go to mobo manufactures page , for my mobo it was MSI Pro z790 A Max and check under support / download you'll find the bios and " click here for a video on how to update bios " msi made a video by themselves showing all steps , just follow the steps and it's done
After bios update only things I did was -
1. Turn off multicore enhancement ( called enhanced turbo on MSI )
2. Change cpu lite load to whatever number you want , mine was at 20 as default so i brought it to 10
3. If everything is stable after changing above stuff run some tests or games you play regularly and check temperatures, performance and voltage and then you can start applying negative voltage offset in bios slowly lowering voltage till you find the perfect spot for your CPU and you're done.
@@Supernova094 thanks man
I have had a 13th gen for a while and in the last ~5 months been experiencing random crashes. I just learned about the voltage issues a week or so ago and have been wondering exactly this. Is my CPU too far gone? Thank you so much for this video!
One they start it’s already damaged irreversibly, see if you can RMA or warranty because intel has been really slimy about thai
First Descendant shader compiling during start-up is a good stress test too.
If it fails and the game crashes, you got problems.
If it survives and the game launches, you're good.
its not strong enough
ya already download n play.. seem like my cpu good to go
Just upgrading the bios can be risky if you are on the MSI board. The first thing you have to do after that is checking AC/DC loadline either in HWinfo or in the bios. It shouldn't be higher than 1,1 mOhm. But MSI is pushing 1,7 mOhom for most of their boards recently which causes EXTREME overvolting.
The easiest noob friendly fix: go to the BIOS and set Lite load mode to 12 (it should be at 0,8/0,8 mOhm). This way you don't have to change LLC from auto and keep CEP enabled without loosing performance.
You can also change AC/DC LL manually but I don't recommend messing with that if you don't know what you are doing.
1:40 this is not entirely accurate information. YES Chips do degrade, however under 'normal' use that degradation takes HUNDREDS of years. Your CPU and GPU will live FAR longer than YOU ever will. It is only under excessive heat and voltage that the degradation is accelerated. With proper cooling your components will last literally forever.
Basically what he said
Damage control ! He needs to cater to Intel viewers ! Didn't see the same "caution" last year with AMD thou !
More like 5 to 10 years for cpu of this century. That you double if you use you cpu only during the day.
Litographie effect can reverse easily in super thin transistor
@@nab-v1w You high ?
Yeah, but like it's fair to say "it doesn't degrade under normal circumstances" cuz the process is too slow to take into account for an user who'll use a cpu for like 5-10 years
My 13900K was bad from day 1. I must have gotten one of the oxidation ones as I got mine in Nov 2022. Consistently crashes when compiling shaders for certain games. Particularly those using UE. Otherwise, it works fine for older games that don't use 100% GPU. Starfield, though published after CPU release, uses an older engine, which works fine. Fails DLSS test on 3d Mark. RMA'd 4090 twice. Thankfully, you figured out it was the CPU. Had no trouble getting RMA as I had a boxed CPU since I built PC from scratch with parts from Amazon. And your build videos to put it together. Waiting for new one to come tomorrow as I got advanced RMA to eliminate downtime.
Thanks for tackling this, Hardware Unboxed and others won't touch the microcode updates or any of this stuff. They say their not AMD fanboys, but they won't lift a finger to support Intel users, so, I think they are.
Bro, i dont think these channels (maybe except linus) do that. Gamers nexus is waiting on the failure analysis lab reports and has mentioned they have a multi part intel coverage coming. HUB has been calling these amd cpus a waste, and made a 30 min video of amd shooting itself in the foot with bad products etc. Comments, on the other hand... i would 100% agree. Its ridiculous how much fanboyism there is
The tests in these videos shouldn't be relied on. Jay is just wrong on some things unfortunately.
@@ZenAndPsychedelicHealingCenter He said it was a series of tests to " Hopefully " determine the state of your CPU. The video was more informative than what you posted.
With my 13900k and 14900k, I have never been able to run Cinebench R23 multicore without several cores hitting 100C. This is with an Arctic Freezer II 420 AIO with 6 fans in push/pull orientation. I am blown away by the temp you are showing there. I've tried remounting and repasting, worried it was a contact issue, but nothing has changed. Both boards I am running are MSI. One is Z790 MEG Edge WiFi with DDR5, and the other is an MSI Z790 Gaming WiFi.
Did u use contact frame?
I got some friends looking for a pc and they keep sending me intel builds, I cant recommend anyone getting an intel cpu for a while till i'm sure it's ok.
Jay!! I want to leave here my feedback! Running a 13900KF, on a Strix Z690-G. After microcode 0x129, I was getting max VID at 1.50 to 1.52v with Intel Default Profile, MCE off, Intel Fail Safe VID bahavior and enforced power limits, getting 35K in R23. I have tried the Asus profile, with MCE on, Auto VID behavior and disabled power limits and for my surprise, im getting 1.42v max 😱 and better Cinebench R23 scores at around 40K. No manually tuning. Never had any issues with my cpu in almost 2 years of use, but got worried about this voltages and degradation issue and updated to this microcode
Hey Jay, great video as always!
However, I noticed that you’re using VID to show the voltage sent to the CPU. Am I wrong, or shouldn’t you be checking the Vcore value, which is listed a few lines below in the HWinfo? From what I understand, VID represents the voltage requested by the CPU, while Vcore is the actual voltage delivered to the CPU by the voltage regulator, and these values aren’t always the same. Please correct me if I’m wrong!
VERY helpful video. Thank you Jay and team!
Think Jay needs to do some more research on this topic. I have a Asus Z690 and Asus profile undevolts my 13700k at base speeds 5.3mhz. The default Intel profile OVER VOLTS. This cpu will overheat nearly instantly on the intel profile and can barely hold 200 watts because it's running at 1.55 volts. The Asus profile under volts at 1.45 and can sustain 253 watts on any benchmark. I consistently get ~30500 pts on R23 with the Asus profile, while the intel can't even muster ~27000 because it's overheating so badly. Not sure what processor Jay is seeing Asus is overvolting on as the default Intel labeled settings are what are over volting. I'm personally not having any issues, tho I know the 13700k is not as affected by this issue, but it seems some are. I just want to know what CPU and MB Jay is talking about instead of just blanket calling out a brand. As I wonder if he understands which combos he's referring to. And yes this is on the most recent bios 3802. But this behavior is the same as its always been, even before the microcode update as I got mine in Mar 2023 and it's functioned this way since the start.
Same CPU and MSI 790 motherboard and this mirrors my experience as well. Think I'm getting around 29,700 but otherwise identical results.
The safe voltage is 1.3 but it reduces performance.
Intel won't force you to get to these levels otherwise they will have falsely advertised.
Instead the microcode just prolongs the CPU enough to get past warranty period.
Thought it was well known the P cores are the problem.
@@McFamily86 that's good to hear. You'll might crack 30k by making sure nothing is running in the background when you test it or changing the priority to high in task manager. But high 29k score should mean you're likely just barely over saturating your cooling. I'm using the Arctic 280 AIO. There's so much confusion with this topic. Sadly Jay is adding to it imo.
@@glyntaylor-williams3593 Yea this is not a MB problem in my opinion. Sure they didn't help. But Intel is straight up lying about the performance and volt requirements. I've even gone as far as to just set my 13700 to the 90 degree max. I only expect to keep it for 5ish yrs as it should last that long, and be powerful enough to meet my needs for that time period. When I bought it, I had all the expectations in the world it would easily last that long and that cooling would not be a major issue. Boy was I wrong. So far I have no crashing issues, I did at the start of ownership of it however. As figuring out the correct voltages was a nightmare because everyone's' "use this xyz" setting did not work. The variation on the 13/14 gen in what works is wild. Which really re-enforces the it's not the MB's causing these problems. It's Intel.
Final thought, I also have 2 P cores that LOVE to blast right to 100c on the simplest of workloads. They are not the prefered cores either. Core 3 and 7 for me love to light themselves on fire at the first chance they get. This does worry me about my chips long term survivability. Time will tell. Likely when it's out of warranty....
Dude 1.45v sucks for a 13700k. You board or CPU are Trash. RMA it. Mine runs a 1.22v because Asrock is much better than Asus.
I finally found my answer and to this day i have 0 issues with my chip!
13500 is based on the Alder Lake die. It should not suffer from degradation issues.
I know it's cool to hate on ASUS right now, but ASUS along with Gigabyte both significantly undervolt the CPU's with their default settings vs Intel settings due to the setting of the AC LL set to 0.4 mohms. In my testing, the "default Intel settings" is what overvolts my CPU the most out of anything right now. In both ASUS and Gigabyte BIOSes, you can manually set a voltage limit so you can stop it from going over whatever voltage you consider safe. I personally don't want mine hitting 1.55v which is Intel's limit with the new microcode, so I set my max to 1.4v or 1400mv. MSI refuses to add this option to their boards. Bottom line, if you know what you are doing, NOT using the Intel Baseline Profile can be better than using it.
Why some RUclipsrs are ignoring this?
that's the 125 bios update?
@@MaxIronsThird Nope, 129.
@@AthosRac No idea, guess not willing to do the research or work to figure it out. I've seen so much misinformation spread.
Do you know if asrock boards let you set those limits?
Man I feel better or a little more at ease now as my 13900KS. It performed pretty close to your CPU. I was not having the problems that everyone else was having and did not experience any crashes before all this hoopla so I think I got really lucky with my CPU (knock on wood, lol). I ran the Heaven, the R23 and I have a Nvidia RTX 4070ti and always install the newest drivers and have never crashed on installing those, so I think I'm good. (knock on wood again, lol). I also have an MSI Z790 MEG ACE MB with the latest BIOS and ME so thank you for doing all this with an MSI BIOS and not ASUS. Thanks Jay for making this Video!👍
I wonder if the Intel problem has to do with the chips bending, because I have a contact plate installed on my 13900K and have never seen any of these issues.
Had contact frame on my 13700k from day one and it still became unstable, so likely unrelated
@@Preemz Ah okay!
@@PreemzI had a ThermalRight contact frame installed and recently a lot of tweaking undervolting and underclocking made my CPU displaying BSODs when running cinebench R23 then I started trouble shooting by removing Ram 1 by 1. The test was passing without bsod with only 1 ram installed then there was last hope I removed the contact frame and I ran the test several times and also stressed for 12 mis. No sign of BSOD. I don't know what I did wrong whether I overtightened the screws or may not tighten enough but I suggest you try it with Intel's stock ILM maybe you don't get BSOD again and do let me know.
@@gameotic1I was very careful to not overtighten the contact frame when first installing, and was very careful to apply even pressure across the IHS with quarter turns. Anyway, I've spent enough time troubleshooting Intel's mistakes already. I've also changed platforms over to AMD this past weekend, and am in the middle of refunding the chip. I'll sell the rest of the platform locally and just eat whatever the remaining cost is.
@@Preemz I've watched Gamers Nexus and other tech tubers guide and installed two of contact frames one for my own system and one for brother's both were working fine until last week my setup after lot of tweaking shown up this issue but my brother's system is still running fine. Maybe I haven't tightened the screws for my system and this problem emerged but I don't think it's necessary because a lot of people are using stock ILM and it won't bend the CPU really easily. Also with Intel's stock ILM the thermal paste distribution was very even so I don't think I need that thermal right contact frame again...
I have MSI MEG Z790 ACE mb with 13900k from 2022 week 46 and its been rock solid. I didn't even bother to turn off XMP and it passes IBT single and all cores. When it dies I'll get something else but it is under warranty now until 2028 per the intel warranty checker. The 2nd MSI 0x129 bios 1.C0 won't allow the CPU core voltage to run up like the previous ones did. FWIW Intel based systems have always been more stable for me than AMD and I've been building my own for over 30 years.
This is why I miss Intel OEM boards. You wouldn't have this problem with one that is calibrated for the processor properly in the first place and not monkied with by default.
Intel definitely have reference boards, they use them for internal V&V / release testing. I am super curious (1) why Intel does not sell them like they used to, and (2) why the internal testing did not catch any of these issues. Voltage spikes were clearly seen, even if they are on a non-degraded CPU
Bring back the BNC connector and SIMM memory!!! LIFE WAS EASIER!!!!
Thank You, Jay! Cheers!
Ok someone feel free to correct me here but heres my ideology. What I think a intel microcode is like the cpu's own "default bios settings" that it will solely abide by. Then theres the UEFI where it intefaces with the cpu to deliver correct power and do all ur overclocking and stuffs. What I assume is that ur UEFI has way more power than ur CPU's Microcode therefore if you manually tune everything to low voltages in the UEFI, it would override the microcode "default bios settings". So Im currently thinking I dont need to update my bios at all since I already tuned my cpu in the UEFI to always consume 1.25v no more no less with stock clocks and hyperthreadding disabled. There has been no degredation and I haven't been getting any WHEA errors in event viewer. Someone feel free to correct me here!!!!!!!!!!!
Jay being real and helpful.
I’ve noticed some other channels who were screaming about this problem have been silent since the microcode update. They slammed intel as the sole problem and said the chip’s performance would have to come way down to fix the problem.
So now we know:
-Intel was at fault.
-The board partners were also pushing the cpu like they shouldn’t have been.
-Intel should replace bad chips and extended the warranty.
-the micro code doesn’t degrade performance much if at all.
-Jay has been the most measured on this (IMO). He has been able to do testing, etc, but from many others not a peep.
Is that because there’s not much new to scream about any more ?
Btw I run a 5800x3d not intel.
Degradation has always been a thing if you pushed your CPU hard enough:
My 2600K from 2011, could original do:
5050MHz without HT
4850MHz with HT
4 years later, 2015, it had degraded to:
4850MHz without HT
4700MHz with HT
8 years later, 2019, it had degarded to:
4700MHz without HT
4600Mhz with HT
11 years, 2022, I gave it to my nephew and it has degraded to:
4600MHz without HT, which is how he has been using it for a couple of years now
4500Mhz with HT
What if it's the motherboard 😅.
And my 2600 K is running fine. The issue is you are running it outside its design spec. If you didn't, it wouldn't have degraded. It's still running well, well beyond its design limits. CPUs will not degrade if run within their design specs - within any reasonable limits that a normal user will see. What kind of voltages were you giving it to get that 2600K running at 5ghz? Just like the issue with 13/14th gen, you were running it beyond its voltage design limits. Then it will degrade. Jay's allegation that there's a "ticking clock" is not true, unless you're pushing it well outside its envelope. You/ve got a ~135% overclock, even after 11 years of abuse. The fact that 13th/14th gens were degrading in *months* is another story entirely.
Also, your degradation is -likely- not just the CPU. Your electrolytic caps on your motherboard are drying out, and the board is becoming noisier. I would be surprised if a re-cap didn't bring you back up somewhere closer to where it was new. Capacitors and other passives on the board are way, way likely to "degrade" well before any CPU.
@@christodd3361 You are right and it still running stable at 4600Mhz without HT, so not complaining 🙂I always ran it 100-150Mhz below the maximum OC and never had any stability problems with it. I have not consider that it could be the motherboard caps and it has been running in the same ASUS P8Z68 DELUXE motherboard for 12+ years, so maybe it is the motherboard and not the CPU. For voltage I used voltage offset for the CPU and believe the maximum voltage was 1.45 V, although it rarely hit that in games.
To say check if your cpu is ready for RMA is carzy!!
Not just CPUs but any and all new hardware should outlast all upgrade paths and replacements, and selling it used is part of the cpu value that intel has denied its customers.
I highly recommend closing Chrome before running test. It can eat up CPU resources especially if there's a bunch of tabs open.
Update: Ideally, you shouldn't have anything running to the best of your ability. Obvious things like Windows services are fine, but avoid running things like torrent downloads, Steam downloads, your browser, or anything similar. This will give you a Safe Mode-like experience to test the CPU.
i think people only kept their PC's on because powering up and powering down might be detrimental to certain components.. like mechanical harddrives. but there are fewer and fewer mechanical harddrives being used these days. people might have 1 large one just for general purpose stuff and things that dont need fast loading.
Not just that, if you live in a colder climate on and off causes stress on the motherboard and other components expansion and contraction. A pc left on is generally more reliable. It's like a car that does long distance start stop twice a day vs short distance with 10x start stop.
A desktop CPU should never degrade to the point of noticeable failure within the useful lifetime of that processor. I still have a Core2Quad Q6600 that works fine, and that CPU spent it's whole life running a 40% overclock!
I have an i5 2500k that ran 1.45v 4 something ghz for ~2 years (I wish I remembered that cpus clock speed) and a Xeon X5670 that ran 1.5v 4.7ghz also for ~2 years. Both are on a shelf in my room, fully functional still.
Granted who knows if either degraded faster than normal since they were only in service a couple years, but at the very least they hadn't degraded enough to need loss clock or more voltage
@@BigMan7o0 and my laptop 13980hx hits 1.41 with no overclock. Jesus, what's intel thinking.
@@Codyslx yeah I've been seeing those voltages one newer Intel cpus and always felt like it HAD to be way too high at full load for a stock config, but ive been on Ryzen for probably 4 years maybe even 5 now so I thought maybe I had just lost touch with Intel stuff and that they truly could take that voltage and heat 24/7 these days. Clearly not and there was actually a reason for concern xD
You don't need to OC computer components anymore they are so fast now out of the box that OC is a waste just throwing more heat at it for no real gains and in gaming its just dumb to overclock for 5 fps gains or even less specially if you have a 4090 GPU its beast at default speeds why OC it. For a basic user that just games on there computer OC anything is a waste minimal gains for way more heat and power use is not worth it imo.
Fully agree. I do the same, PLUS, I use the best cooling components that I can use. Cold electronic components last longer & are more stable than those subjected to an ongoing blast furnace of heat. Sort of a trait that I picked up from working in large mainframe computer shops long, long ago....
Undervolting with overclocking is the most valuable, but it takes a little time to setup up properly.
None of these will surely tell you if you have a CPU problem. Apps crashed constantly on my 13900K and OCCT didn't show any errors on 3 different CPU tests (1 hours each). Also, other CPU benchmark tools didn't crash. But seller confirmed CPU degradation, and Intel confirmed it. Now waiting for a new CPU to be delivered (was under warranty).
Intel used to be so good many many years ago ... and now this
Like most companies when the suits and useless execs start thinking they are smarter then the engineers
I bought an Omen 45l with i7-14700k and it was DOA. I returned that and bought an Omen 40l with i9-13900KF . After updating it worked one day and then black screened. Totally unresponsive to any key commands. Unit returned today. So scared to even try to purchase another gaming pc. Thank you for sharing how to check your cpu. I'm definitely not buying another Intel product.
Still waiting for Gigabyte to release a non-beta bios.
For me the problem on my 14900k became apparent when I was building Docker containers. RMA is in progress.
As much as Intel is telling everyone to do a Bios update to fix the issue, there are millions of users that are not comfortable or do not have the slightest idea of how to do or want to take the chance of bricking the PC doing a bios update. I don't feel like that is a viable solution to the issue for 90% of the people that have a 13th or 14th gen Intel CPU. Even as somone who has built all my own systems, I won't do a bios update unless there is a way to flip back if something goes wrong. I have a Lenvo Laptop that is now a replacement due to it locking up when doing a bios update using the Lenovo updater that was installed and kept notifyingme of the update everytime I turned it on, so I did it and it locked up mid update and bricked itself.
day 1 a week ago. Date of purchase?
What is the exact issue?
How long have you been using the system?
Is the system working before?
Were there any recent changes made to the system, such as hardware or software before the issue occurred?
What troubleshooting steps have you tried to isolate the issue?
Are there any signs of physical damage in the unit?
Did you overclock the CPU?
What is the exact motherboard model used to run the system?
My brand new 14700K started with 0x129 bios :)
My 14700KF temp wreaking havoc rn I can't tell if maybe just my case doesn't have as good of an airflow or there's something wrong with my aio but I had to disable turbo to stay at extremely low temp cus its keeping my clock speed really slow too (but no effect to gaming since i get high fps on most games even with no turbo) but the 0x129 bios update essentially did nothing for me, sure maybe no crashes but a cpu thats always heavy cooling dependent i wished i had known before i got this haha
@@Lawfyy My current cooler is Arctic e34 duo 😂 60-90c in games, i just order AIO 😁
@@Lawfyy What cooler do you have?
@@Lawfyyother than the obvious physical checks its probable that your specific intel defaults spec dosent undervolt automatically when posible its one of the reassons why i still use the non default presets is because they auto undervolt and get better performance for lower temps howerver obviously this can vary a lot across motherboard brands but gigabyte has been good for me during this insanely overexagerated chaos
I purchased my CPU in January, died in June and took my motherboard with it. I now have a new MSI board and CPU, it took 2 months to complete my RMA's with everything going on.
A positive note the latest BETA bios from MSI with the microcode update, has STABILIZED my CPU FINALLY! Prior I had to underclock my CPU at a ratio of 6! In order for it to work properly.
**laughs in 4790k**
Pls don't laugh at my misfortune :(
my and my dad are still using those.
admittedly, these days im running mine as my buddy's game server instead of my main pc, but still.
I just upgraded from a 2500k to this bs 😒😒😒
Laughs with a 5800x3d ( why am I here).
Ahh I love the irony, I still have a 4790k myself, replaced it with a 13700k back in 2022 since it was starting to show its age playing newer games and trying to record at the same time.
This whole ordeal was a slap on the face, I did got it replaced but it made me re think why did I even upgraded lol
I must have won the undervolting CPU lottery because I've been running my I5 13600k with CPU Lite Lode set to 5 on an MSI Pro A Z690 MB for the last year with no stability problems at all and that includes Ycruncher, Cinebench and Prime95. In CinebenchR23 it pulls 153w with all cores at max turbo and temps never go over around 65~70c
did you change anything else such as cpu core voltage? some on here recommend to set it to 1.25v with llc mode 3, or 1.3v with llc mode 5, but both may be higher temps when idle
@@KnightmareUSA No just "CPU lite lode" mode (which is MSI's "one stop easy undervolting" feature and disabling I.A CEP (which used to default to off but now defaults to on in the latest MSI bios). If you don't disable CEP it and the core voltage gets "too low" then the CPU will start clock stretching as a measure against crashing, however this is very over sensitive and tends to clock stretch well before it really needs to.
You can do a little better with trial and error and adjusting various settings manually but the lite lode setting is basically a table of settings controlled by one "master" value and tweaked by MSI to get good results. With this set to 5 on my i5 13600K CPU core voltage is typically very low and never goes above 1.2v, idle typically is around 0.8v ~ 0.9v
@@atomicskull6405 I have the carbon z690 in which cinebench and prime would reach high temperatures without using one of the settings I mention. The difference is that when idle the system may have a slightly higher temp by about a degree or two, but it's very hard to tell as the room temp varies. Setting to a constant 1.3v perhaps might also annoy some who wish to save energy and prevent wear. I'll try resetting the bios settings and trying different lite lode settings but I suspect the cpu temps will still remain high. I think the high temps when stressed might also relate more to the motherboard than just it's bios, although when in normal use and not testing the temps are fairly low (in the 40s for gpu, between 40-50s for cpu)
that one guy who is running an i9-14900KS with 192GB of DDR5-8800 and 2 RTX 4090s:
18:00 You can adjust the time of the test with OCCT for personal use. The timer is over the the CPU+RAM test.
Just got my warranty approval, they are going to refund me since they said they have no stock for the i9 14900k.
Waiting on the warranty approval as well where they refund my i9 13900K ill be moving to the 19 12900k
Man i just learned like two days ago that my undervolt hasn't been running for maybe a year... no wonder my laptop was getting so hot especially during the summer lmao. Thankfully turning all the vm stuff off fixed that and my new -175mv core -70mv cache undervolt is surprisingly stable.
Updated to the new Asus BIOS with the beta microcode the other day. Was being sparing with my PC (13900K) since this all came out. Using Intel limits and I started getting crashes with Diablo 4 after playing for close to 10 minutes. The try to boot up The First Descendant and I get a “Out of VRAM” crash. Welp here we go…
Were you getting errors and crashes before the BIOS microcode update?
I had similar issues when I upgraded to the new BIOS. Check if your DRAM voltage is sitting on where your RAM manufacturer has tested the RAM to be stable (1.35 V for example, just find your exact RAM DIMM model and see on the manufacturers page, this only applies if you are using OC RAM ofc). And check if your Vcore voltage is potentially sitting too far below your VID voltages. If the temps are fine on your system you could just - for trial and error - go and set a slight OVERvolt (offset mode) to your Vcore voltage, something like 0.01 or 0.02 V while keeping the core temps in check. And do try the OCCT variable CPU + RAM stress testing, its very handy for checking if your system is stable. And of course keep playing Diablo to see if the crashes disappear.
I have an i913900K and ASUS motherboard. I'm not very computer savvy...
The new non-beta BIOS update (version 1663 including the 0x129) from ASUS says "for non-K processors"... and I have a K processor. Do I need to update to the version 1663 BIOS?
Last month I updated to the older 1661 version which states: "Updated with microcode 0x125 to ensure eTVB operates within Intel specifications." Never had any crashes or trouble.
I'm a little confused about the "for non-K processors".
@@turtle2720 I think it just adds the microcode for the non k processors as well, it used to be only for the k processors. so i do´nt think it would change anything for you but it probably wont hurt to install it anyways.
@jeremyashcraft2053 I was not funny enough. Everything was stable and fine about a month ago which was my last big play session. Maybe something in the new microcode and BIOS isn’t playing well together.
I’m going to downgrade my system to a 12900K for now and wait for this to all blow over.