As an engineer at AMD I can say we do not release products pushed to their limits. One of my jobs is finding the limits of the parts I help design. Releasing hardware at it limit or past its limit is not good for customers!
Ive been running cpus at their absolute limit starting with athlon xp and pentium 4... They still work. Also have core2 quad and phenom 2 running at their absolute limit.... they still work fine too. Also all the parts around these cpus have been running at its absolute limit. My i5 12600k is running at its absolute limit for the last 3 years and stable as a dream. Little bit of maintenance goes a long way. Running it at its limit doesn't mean it has to degrade or break. Well at least it didn't in the past. So no need to be defensive about running parts hard.
@@0MeALot0they weren't being defensive, they stated they dont release products at the absolute limit, which is good for the brand image and for people who might not know how it could negatively affect them by pushing it to the said limit
Totally disagree regarding AMD CPU’s. They do not produce anywhere near the heat Intel produces. Intel will degrade over time because of their insane power draw and heat production. AMD doesn’t have that problem. They may one day when they’ve outgrown the present architecture, but Intel has been there for a few years now.
I have a R9 7950X myself and I must confirm they DO degrade over time but the problem is not as bad / slower than Intel but it IS there. Keep the max cores temps below 89 to prolong the life of your R9. The hotter they get the faster they degrade.
@@betawing1914 Asus, MSI maybe Gigabyte mobo's? Got an Asrock X470 Master SLI with a 3600 cpu and have no isuues whatsoever. I've never seen my pbo pushing that far. 1.35v is what pbo sets it to so it is more the issue with the mobo manufacturers then the cpu's. It's not the f-up with Asus, MSI and Gigabyte with their mobo's and 7800x3d burning to a cinder and AMD quickly acted upon that so now every AM5, and probably AM4, mobo sticks to the settings/specs told by AMD. So to come back at what you are saying is BS and you know it.
My experience is actually motherboards failing before the CPU does, recently, something to do with the VRM from what I can tell. Always had this problem, even 10-15 years back, perhaps it because CPU's these days are starting to get really hot again, just like the Pentium4 days.
My 9900k will probably still be going strong long after my 13700kf dies. And the 9900k survived some fairly big overvoltage due to early Gigabyte BIOSs being crap.
I got an 12900kf for $230, could've gotten an 13700kf but the stability issues are making everyone avoid 13th and 14th like a plague, I'd rather have a good cpu that is stable and lives for 6+ years than having a slightly faster one that will die in 3 years even if you downclock it and with the downclock you're probably just at 12900k performance which just plain sucks.
this doesn't make sense. I also saw a video in which they mentioned that server components based on the 13th and 14th gen intel chips were failing at an alarming rate as well. that's undervolted chips specific for stability.
If AMD had any issues we would know by now. This is just stupid. Also, their processors are crazy efficient. The 7800x3d runs at 120 watts. That's crazy efficient for what that processor delivers
7800X3D doesn't run at 120W. Yes, official TDP is 120W and max allowed power is 162W(ppt). However it can't get nowhere close because thermal/voltage constraints due vcache. Stock TDP/power limits are pretty much irrelevant. 120W TDP is actually very high for single CCD ryzen, it's at edge of what's physically possible for such a small die even without vcache. Direct die cooling is pretty much required to get there. In reality at worst 7800X3D does 90W under heavy workloads. Gaming around 60-70W, in lighter games even less. You can set PPT to 90W and never see a difference in any benchmark.
Since when Intel's issues = AMD's issues? They both have different architecture, different thermal goals, different safe-guards, different performance on lower power vs. high power, different TDPs, different "actual" power-draws, not to mention only one company has a track record of CPU issues whereas there other is practically scarce. Good thing it's his opinion because it's a pretty emotional and un-factual opinion. The only thing I agree with here is that the companies need to find a way to get their CPU's to operate underload at lower temps ... AMD at least is trying to do that with Zen 5 presumably operating at 7°C lower than Zen 4 at the same power.
I agree. The 1-2 core boost for limited time is stupid. All CPUs should just be set at the rated all core speeds. Chasing niche benchmarks is pointless. Low quality silicon with very high VIDs are going to over volt and degrade over time, this is inevitable for all CPUs.
Intel needs to at least stop selling the 13/14900k skews and save what production they have left for warranty replacement. There is still the chance this reaches the point of involuntary recall or class action lawsuit, etc.
Totally disagree about Ryzen 3000 and 5000 Series. I have R5 3600 and R5 5600X when they released globally. They did not show any problem and my 3600 is about 5 years. You should have research that 3000 and 5000 series never boost over 5GHz and the voltage still lower than 1.5 volt which are safe.
Agree with current generations of CPUs being pushed way beyond the most efficient operating point. One of the best features of current AM5 AMD CPU / motherboard / BIOS is the ability to cap the max boost clock. Previous AM4 systems can't do that. I have manually limited my 7950X boost ceiling to 5.3Ghz for CCD0 and 5.1Ghz for CCD1. With a slight under-voltage, the maximum voltage the CPU will ever see is 1.15v, which is way, way more comfortable than the default 1.48v. I'd gladily trade 6% single core performance for a CPU that runs cool, quite and for a long life any day. Single core benchmark is lower sure, but who cares. Pumping 1.5v into 5nm silicon just sounds wrong. Enough is enough for me to keep this consumerism going and upgrading computer every year like it's fashion item. If a CPU can't work reliably for at least 10 years it's a failure to me.
The voltages seem to have little or nothing to do with the Intel failures. They are failing at a 50% rate even in server boards that are very conservative in their voltages. Please provide any source that indicates that AMD CPUs are failing.
@@TheArcticDen you mean the Asus boards blowing out the chips that was quickly addressed by AMD with a full replacement and a strong side eye at Asus to cool off that is no longer happening?
So where have you heard AMD chips failing? Do you base it on actual facts or is it just "your opinion"? 95 degrees is within spec and it doesn't affect performance or longevity, so i don't really know what you are on about.. You are basically saying all chips will fail which is an L take, and seems you think all cpu will fail because intel chips has issues atm as if it was an indication of anything omg lol.
Electronic engineers know all electronics die eventually. That is what the classic bathtub curve of failure rate is all about. Bad products have much shallower Constant Failure Rate regions in their curves, like the Intel 13th and 14th gen top end processors are having now.
There's big time diminishing returns when you're at the bleeding edge. I think a good rule of thumb is to limit the hottest core to no higher than 80c and you will probably be fine, assuming you don't have one of the defective intel chips.
He means Intel is a sinking ship. AMD will take the Nvidia position in the CPU market and capitalize hard in consumers because Intel is behind and sinking. We had a good run of competition with zen but.. it's going to come to another decade of stagnation with AMD.
@@christophermullins7163 exactly. If Intel doesn't do something quickly, we'll all end up in situation like GPUs with $2000 price tags because there's no competition.
Sure, Silicon degradation is a thing. However, it should take YEARS for this to happen. Given 13th & 14th Gen are having stability and death issues already.....those will NOT be saught after in the future, and 2nd hand buyers will avoid at all cost! AMD does NOT have this issue, especially with Curve Optimization.
there is a reason why GAME DATA SERVERS now replacing their intel CPUs and replacing them with AMD CPUs and here you have the answer, if AMD CPUs have the same problems as Intel
I think there is something to the concerns with AMD, but this is mainly to do with the X3D chips. The Chiplets issues with the 5800X3D and default BIOS settings were an issue a year ago where they ran into some serious problems but it was only a mild discussion point any no one was actually panicking. I switched from the 5800X3D to 5900x and s=decided to watch and wait on the more recent offerings.
I would have to disagree with this AMD has stated they find the max settings of their CPUs and then purposely turn them down a little bit to prevent the type of problems intel is running into right now even if it will cost them not being number 1 in the benchmarks they believe their customers are better served this way, and after RMAing my 13900k last week i would agree
Havent done any real digging on my 7600x performance like this, but i do know that when i game, its rare for it to exceed 75C, let alone getting up past 90
I'm glad I haven't upgraded my 12900k, I have a z790 and was thinking about dropping in a 13900 or even jumping to the 14900 last month but saw all the issues popping up and decided to hold off until more info came out. So far my 12900k is stable OC'd and nothing weird happening.
Just tead the question posted for the video, " will all intel and amd cpus degrade over time?" My answer is question based. Will a mechanical engine fail overtime? Do spark plug wires degrade over time? If you answered yes, then you understand, nothing last forever.
For a long time I have felt the same way about where products are going...really I don't want to see temperatures above 85 degrees Celsius whether that be CPUs, or GPUs (7900xtx cough cough)...while most of the time the 95c temps come with short term heavy workloads like cinebench which really are not meant to be sustained for longer periods of time...I am in 100% agreement that everyone needs to dial things back to the long term safe zone. To put it in perspective PCI-E 5.0 M.2 NVMe storage drives will throttle starts around 81c...that's with a huge night and day difference in total drive power draw. DDR5 has a max operating temperature of 95c...so just imagine this theoretical scenario... the CPU next door is at 95c, the M.2 is at 81C...and the memory on the gpu is hitting 110c...bad stuff will happen to those DDR5 memory modules running at 1.4 volts with tight EXPO or XMP timings/subtimings. All components are starting to run hotter...and that confluence could very well lead to premature degradation, again they should just set the limit line at 85C for CPUs and GPUs.
both companies let two cores boost beyond what the silicon can handle over time, they most likely do it because we consumers are dumb enough to like the higher numbers on the charts of techtubers when they do game tests without considering that those numbers are achieved by detrimental clocks. Most that have locked the cores to disable the boosting from day 1 have zero issues with their cpu's. The sad thing is that if one finds out too late then the damage could already have been done and there is no saving the cpu.
I bought a used 5900X some days ago and I set the PPT to 105W (from 140W), the max temp limit to 85°C (from 90°C) and the curve optimizer to -30 all cores. In the end I lost a couple hundreds points in cinebench multicore, with the turbo still pushing to 4.950 GHz. Waiting for AM6 for the upgrade.
@@gianlucapx the used price of the 5900X makes my 175$ purchase for a new 5700X early this year look silly 😂. Got it running just under the speed of the 5800X with a +200mhz offset and PBO curve optimizer voltage offsets of between -10 and -25 on a per core basis. 😂 Max wattage set to 100 in bios and it's super stable and runs very cool under full load with a peerless assassin cooler. Waiting for Ryzen 12000 to upgrade 😂
Well.. as Wendell pointed out this degradation also happens on stock settings so setting and forgetting ain't a good idea. Check recent Framechasers video. You need to lock cpu boost clocks to something resonable.
This sounds like a clickbait for AMD CPU users. Because if you've ever used one, you know that AMD has placed hard thermal limits on their CPUs. I wouldn't call that "pushing it too far".
First the burning power cable of 4090 i guess. Then this intel oxidation problem. 💀 Amd is not really far away. 🥹 What the hell happening to the tech world.
thats why a limit my voltage in the bios, and PBO takes care of the rest, upon trying to get highest score on benchmarks and core clock when I first got my R9 5900x, I saw it reach 209w and just below 90 C, but I did not let the test run for long, I tuned way down... I get around 4.7 GHz in gaming and I never seen my processor pass 72 C and that was a very hot day in my room... so I'll be upgrading to the new 9000 series probably, and I will do the same thing, limit the voltage and let PBO do the rest, if I lose 2, 3 or 400 MHz on boost clock so be it... this matters for higher level consumers, devs etc... as they push these processors a lot harder than most of us will ever do
You're just making up controversy about AMD based on nothing. Per wendall of Level1tech , AMD accounts for 0.31% of cpu hardware errors. Laptop chips ROUTINELY surpass 100c and it's been that way for quite a long time. Intel's chips have a DESIGN FLAW, exacerbated by the fact they continue to cram ludicrous amounts of cores into an undersized package size while routinely allowing the cpu to pull 320-450w. I'm old enough to remember when pentium 4 was the big thing, and we happened to have a compaq laptop that had a desktop variant of the 3.2 ghz pentium 4. Which was rated at 82w. For 1 core. The barrel jack on the laptop got hot enough that the solder joints connecting it to the motherboard....MELTED CLEAN OFF! Not once. not twice But THREE TIMES. Despite limiting the power profile. The solution was to take the power connecter from an old Compaq scanner keyboard and solder that to the board to put some distance between the solder joints and barrel jack. The only other Intel chip i've owned over the past 30 years, was a core i7 940 on an x58 sabertooth. Never overclocked the thing, yet that thing managed to pull enough power that the VRMs simply crapped out. This isn't a new problem for intel, it's just the most egregious example. AMD cores are pulling between 10.7 -15.8w each under full load, depending on power limits, intel's P cores are pulling 25-35w each. There's no comparison to make here, particularly when every generation of Ryzen chip for the past SEVEN years has had a max temp of 95c (save for the x3D variants that drop the temp to 89c) where as the 13900k launched less than two years ago, the 14900k has been around for TEN MONTHS. The only issue AMD had was due to motherboard manufactures failing to properly limit power profiles, while intel has encouraged board partners to push everything to the max out of the box up until the point their chips began failing in droves....and yet even with more restrained power profiles newly purchased intel chips continue to crap out.
The problem is: the CPU are NOT getting ENOUGH voltages! Motherboard Manufacturers Power AC and DC load-line settings in the Bios are changing the voltages that the CPU gets, thus not providing enough voltages at certain frequencies under high loads. I can run the 12 gen CPU on Load line calibration LLC 1 fine, but the 14 gen needs LLC 8 to achieve higher frequencies because there is too much V droop at lower Load-line settings, and the CPU does not get enough voltages under higher loads.. 14 gen needs more power and voltages under high loads.. So Motherboard manufacturers just needs to fix their Bios settings ... And it is cool that intel lets us change the load-line settings so that we can under-volt the CPU, but if the under-volt is too aggressive crashes, errors, or instability happens. i tested this, and even the "out of video Memory" error was not present when i just was sending a bit more voltage to the CPU, but it was present if i was sending to little voltage. Apps and games were not starting or crashing with too little voltages, but not crashing anymore with a bit more..
I can almost gaurenttee that all cpus build disposable just like all things corperations produce. is part of their design. they do not want products to last too long
I just had a R9 5950X die on me suddenly one day... I've been gaming on computers since 1995 and this is my first and only cpu to die on me. The thing that is odd is that this cpu I have that died lived a very pampered easy life... never overclocked and has been liquid cooled so never hot either. Rough luck I guess....
7800x3d, PBO under/overclocked watercooled system. Had it since Jan. Run it 8+ hours a day. 0 performance degradation or issues. Also 6200 ram on the mem bus. AMD solid so far
@@Giovanni-Giorgio The intel chip failures come faster than the length of use ive ran mine at. Also Im using a cryosheet, so there will be no repastes.
I had no idea about the Intel 13th and 14th cpu problems until my new prebuilt with a i7 14700KF I bought last week started to BSOD several times.. Thank god it had a 30 day return policy.. I'm most likely will never buy an Intel processor ever again. AMD's new processer releasing next month will be a big hit!
Intel CPUs degrade because of HEAT. This is mostly because Intel and motherboard manufacturers pushed voltages and power consumption of CPUs too high for die to transfer and dissipate. AMD doesn't do that. In fact, heat production (that TDP number) out of the box in AMD CPUs is same as it always has been since Athlon X2 (early 2010s) days. This means, AMD's CPUs don't degrade nearly as much and are usable to end of their usable life. I still use those old processors to my daily computing needs and those still work just fine.
The high voltages and click speed theory has been debunked by Lendel at Level 1 Tech. Server class boards running 13 and 14 gen CPUs also have problems and they are built for stability
Nobody knows why Intel CPUs are degrading so quickly. People talk about the over-voltages on the motherboards (which was real for many of these units, including my 12900K), but there are units which have NEVER been overvolted or overclocked, which degrade just as fast. It is 100% speculation at this point, and it is highly improbable that your local internet celebrity / yahoo will have the knowledge, equipment and experience to figure out the actual reason. Hopefully Intel engineers are burning the midnight oil on this, as this seems to be shaping up to be a huge problem.
You know how old school CPUs could last you 20 years if you wanted to? Intel and AMD realized that most people are upgrading every 2-5 years and will never have it that long. So, why not push them harder, knowing they’ll degrade faster? Who cares, they’ll be buying a new one soon anyways. The problem is, I don’t think they expected the degradation to be this bad, this soon. They got caught with their pants down.
What the man says makes sense, and to mention it again, no AMD doesn't seem to be having the same issue as Intel. That doesn't mean they won't ever get the same problem if they continue this how shall I say ummm, pissing match, and it is that, when the makers knowing set limits to high, they willfully are the bad guy, and for what? The title of we're the fastest? Who cares when the CPU is fried, it goes from the fastest to the deadest 😂 Perhaps AMD will learn from this mess, maybe they won't. Let me ask this , why pay a good amount of money for a CPU that's going to kill itself off much sooner than it needs to? Don't be sheep people you don't have to buy the biggest, or fastest anything to achieve a great computer experience.
From level1tech data Intels problems look to be memory corruption which could be cache or Memory controller hence down clocking and dropping to single data rate not double data rate is helping as strain on Memory is reduced and Memory controller has more time to refresh memory cells. I'll speculate It could be Memory controller can't charge Memory up to level for 1 to be maintained and it's dropping to point it's seen as a 0 and data is then corrupted CPU slows as it trys to correct that then crashes which is symptoms Wandel of level1tech was talking about so no software would fix it unless it's to slow cpu down so controller has time to manage Memory.
Just run AMD in eco mode. Problem solved. This is the mode they should be using by default. With Intel? Do you even have an eco mode? I don't know. My pc has never crashed. In 14 years. I push nothing. But I will need to upgrade at some point.
Anyone thinking architecture shrinking brings only positives but what are they thinking? If something is small enough it can be killed with really low voltage and amps if architectures continues to shrink this will happen more often and sooner after buying cpus gpus or whatever so i think architectures started to hitting physical limits even if they are capable to operate at lower voltages and amps and every chip will fail someday that is for sure and anyone did not mind radiation in nature that can also affect lifespan if tranzistors are small enough and exposed to it for a long period of time 20A will be not really affected or measurably at least i think but what about nanosheets or whatever...and i don't talk about pcs at home but in hospitals where are some radioactive isotopes stored and used will they develop some form of shield for pcs?!
@@Hatsunari_Kamado @_mikkoxd_292 this but still, like Wendel said at The Full Nerd Podcast, update your mobo's firmware even if you have a 12th gen Intel CPU.
lol if your CPUs have degraded that means that you ran them stock or let the Mobo auto OC and I thought you're an enthusiast. I've been running my 9900k OCed to 5.1Ghz all core and 4.7Ghz cash at 1.31V with 4000 cl16 RAM for over 4 years and no degradation at all, I'm very happy with the performance and have no plans to upgrade anytime soon but that said don't think it will be to an Intel CPU, probably going Rizen x3d next, especially if it can be OCed. Learn to tune your hardware, it takes 1 time a few hours at most and you're set for years.
We all know that all CPU''s degrade over time its the reason why my 3770K can't even run on default clocks without increasing the voltage after it was running on OC for around decade the issue with Intel right now is that the processors are degrading fast and its on level never seen before AMD doesn't have that issue they have been pushing it for a while now and have been cranking those voltage since Zen2 I have seen multiple motherboards crank up to 1.5v even when idle and those chips are still alive and well to this day I am not saying its not degrading them cause I have had a 3600X that won't work on default clocks without cranking the voltage to 1.4V and it wasn't even overclocked if your someone really worried about degradation then you need to look into under-volting and using the curve optimizer brings more performance with lower temps to boot.
Take a 20-core chip like the i7-14700. With a ring bus, you need 1 set of transceivers per core, for a total of 20. With a mesh, you need (19*20) 380 sets of transceivers AND each core needs extra arbitration logic and buffers for the 19-channel multiplexers. Not to mention that they can etch the ring interconnect on a single layer of silicon, whereas the mesh will need many more layers than that just to route it. The mesh will be somewhat faster during high utilization, but even at peak loads most of the bandwidth will be wasted as its highly unlikely that every core needs to send data to every other core. They probably did a cost-benefit analysis and the answer to using a mesh was probably "OH HELL NO" after only a few minutes of thinking about it.
I expect zen5 will be such a banger of a cpu and playform. Lower default power, better pbo and boosting, amd clearly sandbagging to throw off intel(although.. if i can see it, intel can see it). I expect +20% in games and then a +20% for the vcache version on top of that because they cut the latency of the cache system drastically ao we will benefit even more from the stacked cache. Yall get ready for an absolute bloodbath. Intel will be out here selling "similar performance" cpus for half the price of amd because no one wants them. Heck.. their margins in current cpus are horrible compared to amd so they wont be able to profit at all. Im sure they will come back at some point but just like amd after bulldozer.. Intel will have to rebuild trust and find ways to build more efficient and profitable cpus.
When i was in computer school I saw size of chips schematics If drawn on paper it was size of huge wharehouse They shoukd shave bad cpus see if there hot spot My thoughts there line too close long or bent to sharply 2 nanometer you get em interference so it needs em shielding This prevents hacking and malware My motherboard acted up and toasted my hhd
It is also good that Intel wants to hold unto their semiconductor factories thru placing 20 billion toward their advancements to near atomic sizes. Where node size competition will soon be irrelevant when near atomic node size have been reliably reached in the near future. This also means that once this happens. It will not only allow Intel to produce equivalent quality products at superior prices than their competition can even afford to do for them having to rely upon other companies like TSMC to produce their chips. Thus forcing competition pricing down to a minimum in a world that cries for economic prices. But it will also mean that stability issues will be less likely thru that as well. Especially for more performance with less heat and electricity needed too. Thus Intel's success and holding unto its semiconductor factories will certainly benefit seemingly everyone, save but the Nvidia guy and his hopes to sell everyone graphics cards at nearly 2 to 4 times their value of course, hahaha.
Hmmm, do you member 5Volts 8088's? 80186 (mostly used as MCU not CPU, HDD controllers in particular used this baby many a time) , 80286, 80386 and the big majority of 80486, only with the Pentium family did we start to see lower than 5V core volts as the new normal, specifically 3.3V, we have to go to the Pentium 2/3 to see voltages at 1.x volts. So, you are speculating like there's no tomorrow, that said the CPUs are ageing way too fast if this is not a factory fault (well would still be factory fault, just a different one) it may be a raw metals quality problem to start with, which if true means only the factories receiving lower quality metals would be impacted which would completely invalidate your AMD theory, and yes theory because there are 0 confirmed facts here for it to be anything else.
I am a computer builder i build on ryzen and intel platforms let me remind everybody ryzen 3000 cpus where also dying at a high rate because of pbo pushing over 1.4 volts through the cpus. Nobody remembers this because its amd. Lock down your cores never exceed 1.4 volts and u will be a happy gamer for a long time no matter the cpu
You're wrong. But that's okay, Intel's 10/11 and even 12th gen are fine, the reason why they degrade is down to poor engineering, they moved the I/O layer, which us why this only impacts 13 and 14th gen. Don't forget that click wise there's not much if a difference between a 12900K and a 13900T fir example. Bad engineering is what lead to this. Nothing more
i wonder about the 3d series processors since they run a bit hot on average but everyone I know with one including myself either turns off all "performance boosts" their mother board has and runs them bone stock or under clocks them.
@@asdasdasd-gx7zs i had to lock my voltages, because i use a 3900x. My 5950x was screwed up because of voltages and i had to get rid of it. Can't say anything bad about amd apparently.
If you would do your research instead of parroting you would find most of the fails are on Asus MB. Asus as default pushes 100A beyond the specs. Of course it will fail.
absolutely right i agree we are pushing cpu and motherboards to way fair hince the issues arising now we are trying to get somewere not really needed i think the issues are vendors in all sectors in terms of cpu motherboards and graphics cards needs to sort of work together to come up with a solution on all sides that work inharmony and untill that happends the same issues will arise and yes its easier said than done but if business means as much to them as they cliam then thats the only way to sort and fix also theres a big enough market out there to satisfie the market but really oh iam kinding that wont happen so problems will keep happening untill gives hay its happend many times and will keep going and going till it ends up blowing up in many faces literally and hypothetically thats when they will go whoops we should of gone an other way and by that time too late so let see what happens hopefully they find a way to sort it but cant see it 🤔🤫🤨😒
x86(x86_64) is long death. So engineers need to thinker about new tricks to still get some performance boost over the last gen, so they can still sell something. And both lie and lie and lie about it. You know that you are at the end of the road, when only more cores, more and faster cache comes to rescue. Would be funny to compare each generation benefit on the time an single instruction needs in ns. How many years was needed to go from 1Mhz to 1Ghz? How long from 2Ghz to 4Ghz? And now we fighting the 5Ghz Bullshit war. Amdahl's law kicks in, and apple keeps smiling with there non x86 cpu's.
7950 x 3D 5.8ghz. all core, posted a video on this dilated, liquid metal.... Still 100% stable today. Please consider changing your name to mentally challenged. It would be more accurate.
As an engineer at AMD I can say we do not release products pushed to their limits. One of my jobs is finding the limits of the parts I help design. Releasing hardware at it limit or past its limit is not good for customers!
Ive been running cpus at their absolute limit starting with athlon xp and pentium 4... They still work. Also have core2 quad and phenom 2 running at their absolute limit.... they still work fine too. Also all the parts around these cpus have been running at its absolute limit. My i5 12600k is running at its absolute limit for the last 3 years and stable as a dream. Little bit of maintenance goes a long way. Running it at its limit doesn't mean it has to degrade or break. Well at least it didn't in the past. So no need to be defensive about running parts hard.
@@0MeALot0they weren't being defensive, they stated they dont release products at the absolute limit, which is good for the brand image and for people who might not know how it could negatively affect them by pushing it to the said limit
Totally disagree regarding AMD CPU’s. They do not produce anywhere near the heat Intel produces. Intel will degrade over time because of their insane power draw and heat production. AMD doesn’t have that problem. They may one day when they’ve outgrown the present architecture, but Intel has been there for a few years now.
I have a R9 7950X myself and I must confirm they DO degrade over time but the problem is not as bad / slower than Intel but it IS there. Keep the max cores temps below 89 to prolong the life of your R9. The hotter they get the faster they degrade.
Well actually some mb was actually giving the am5 chips to much voltage resulting in them burning up over time
Your wrong bro ryzen 3000 cpus where dying because of pbo pushing 1.45 volts nobody remembers this because its amd
@@betawing1914 Asus, MSI maybe Gigabyte mobo's? Got an Asrock X470 Master SLI with a 3600 cpu and have no isuues whatsoever. I've never seen my pbo pushing that far. 1.35v is what pbo sets it to so it is more the issue with the mobo manufacturers then the cpu's. It's not the f-up with Asus, MSI and Gigabyte with their mobo's and 7800x3d burning to a cinder and AMD quickly acted upon that so now every AM5, and probably AM4, mobo sticks to the settings/specs told by AMD. So to come back at what you are saying is BS and you know it.
My 3600 died because my dumbass brain turned on PBO and thinking 1.45v is fine. Its biostar x370 btw
Userbenchmark is that you?
I run my i7 4790K from a decade ago every single day, all day. It is overclocked to 4.6GHz ALL-CORES. It isn't degrading whatsoever.
The same with my 13 Year old Bulldozer. 5ghz since day one.
What a beast 😂
It almost certainly is degrading, just not going to be noticeable for a few decades is all.
I still have a 4770k Somewhere , at 4ghz and 10y old,still rolling.
it's the architecture issue with the 13-14th gen i9s apparently
My experience is actually motherboards failing before the CPU does, recently, something to do with the VRM from what I can tell. Always had this problem, even 10-15 years back, perhaps it because CPU's these days are starting to get really hot again, just like the Pentium4 days.
He's doing damage control... didn't you see the merch he got from Intel in the intro?😆
Intel Lemon Lake
Hehe 😂
💀
Chernobyl Lake
"Problems Inside"😅
Lava Lake
I’m still thankful my 9900k and 10850k have been good to me for all these years.
My 9900k will probably still be going strong long after my 13700kf dies. And the 9900k survived some fairly big overvoltage due to early Gigabyte BIOSs being crap.
Yeah I have a 9900k on a Z390 and had to nerf every setting to get that voltage down. Gigabyte was buck wild for that 😂@@stevexc222
i just replaced last year my 10y old 5820k by a 13600k. they doesnt seems concerned yet so im hopeful but worried.
I got an 12900kf for $230, could've gotten an 13700kf but the stability issues are making everyone avoid 13th and 14th like a plague, I'd rather have a good cpu that is stable and lives for 6+ years than having a slightly faster one that will die in 3 years even if you downclock it and with the downclock you're probably just at 12900k performance which just plain sucks.
so to distribue rumors like AMD CPU will degrade like INTEL CPU certainly helps INTEL a lot ?!
this doesn't make sense. I also saw a video in which they mentioned that server components based on the 13th and 14th gen intel chips were failing at an alarming rate as well. that's undervolted chips specific for stability.
If AMD had any issues we would know by now. This is just stupid.
Also, their processors are crazy efficient. The 7800x3d runs at 120 watts. That's crazy efficient for what that processor delivers
7800X3D doesn't run at 120W. Yes, official TDP is 120W and max allowed power is 162W(ppt). However it can't get nowhere close because thermal/voltage constraints due vcache. Stock TDP/power limits are pretty much irrelevant. 120W TDP is actually very high for single CCD ryzen, it's at edge of what's physically possible for such a small die even without vcache. Direct die cooling is pretty much required to get there. In reality at worst 7800X3D does 90W under heavy workloads. Gaming around 60-70W, in lighter games even less. You can set PPT to 90W and never see a difference in any benchmark.
@@kognak6640 and even then it's the fastest gaming CPU on the planet. Even more efficient than I thought then.
Since when Intel's issues = AMD's issues? They both have different architecture, different thermal goals, different safe-guards, different performance on lower power vs. high power, different TDPs, different "actual" power-draws, not to mention only one company has a track record of CPU issues whereas there other is practically scarce.
Good thing it's his opinion because it's a pretty emotional and un-factual opinion.
The only thing I agree with here is that the companies need to find a way to get their CPU's to operate underload at lower temps ... AMD at least is trying to do that with Zen 5 presumably operating at 7°C lower than Zen 4 at the same power.
I agree. The 1-2 core boost for limited time is stupid. All CPUs should just be set at the rated all core speeds. Chasing niche benchmarks is pointless. Low quality silicon with very high VIDs are going to over volt and degrade over time, this is inevitable for all CPUs.
Intel needs to at least stop selling the 13/14900k skews and save what production they have left for warranty replacement. There is still the chance this reaches the point of involuntary recall or class action lawsuit, etc.
Totally disagree about Ryzen 3000 and 5000 Series. I have R5 3600 and R5 5600X when they released globally. They did not show any problem and my 3600 is about 5 years. You should have research that 3000 and 5000 series never boost over 5GHz and the voltage still lower than 1.5 volt which are safe.
Agree with current generations of CPUs being pushed way beyond the most efficient operating point. One of the best features of current AM5 AMD CPU / motherboard / BIOS is the ability to cap the max boost clock. Previous AM4 systems can't do that. I have manually limited my 7950X boost ceiling to 5.3Ghz for CCD0 and 5.1Ghz for CCD1. With a slight under-voltage, the maximum voltage the CPU will ever see is 1.15v, which is way, way more comfortable than the default 1.48v. I'd gladily trade 6% single core performance for a CPU that runs cool, quite and for a long life any day. Single core benchmark is lower sure, but who cares. Pumping 1.5v into 5nm silicon just sounds wrong. Enough is enough for me to keep this consumerism going and upgrading computer every year like it's fashion item. If a CPU can't work reliably for at least 10 years it's a failure to me.
The voltages seem to have little or nothing to do with the Intel failures. They are failing at a 50% rate even in server boards that are very conservative in their voltages. Please provide any source that indicates that AMD CPUs are failing.
5950x had this problem, but was kicked under the rug...
@@TheArcticDen you mean the Asus boards blowing out the chips that was quickly addressed by AMD with a full replacement and a strong side eye at Asus to cool off that is no longer happening?
@@chemislife I think that was for Zen 4 X3D chips, correct me if I'm wrong.
So where have you heard AMD chips failing? Do you base it on actual facts or is it just "your opinion"? 95 degrees is within spec and it doesn't affect performance or longevity, so i don't really know what you are on about.. You are basically saying all chips will fail which is an L take, and seems you think all cpu will fail because intel chips has issues atm as if it was an indication of anything omg lol.
@@NulJern zen3 the 3000 series and 5950x had a voltage issue, granted am5 doesnt seem to have that issue.
Electronic engineers know all electronics die eventually. That is what the classic bathtub curve of failure rate is all about. Bad products have much shallower Constant Failure Rate regions in their curves, like the Intel 13th and 14th gen top end processors are having now.
Covid learned me to be scared of everything. Cpu degradation, oxydation.... I can't go out anymore
Finally somebody actually admits that Intel 12th Gen has issues
I'm glad I'm not the only one with a 12900K that has issues
13 th gen i5........5.7 ghz 1.31 volt .... With a undervolt of 0.035 no hyper threading and runs like a dream
There's big time diminishing returns when you're at the bleeding edge. I think a good rule of thumb is to limit the hottest core to no higher than 80c and you will probably be fine, assuming you don't have one of the defective intel chips.
This is the end of everything 😢
Wdym
He means Intel is a sinking ship. AMD will take the Nvidia position in the CPU market and capitalize hard in consumers because Intel is behind and sinking. We had a good run of competition with zen but.. it's going to come to another decade of stagnation with AMD.
@@christophermullins7163 exactly. If Intel doesn't do something quickly, we'll all end up in situation like GPUs with $2000 price tags because there's no competition.
Sure, Silicon degradation is a thing. However, it should take YEARS for this to happen. Given 13th & 14th Gen are having stability and death issues already.....those will NOT be saught after in the future, and 2nd hand buyers will avoid at all cost! AMD does NOT have this issue, especially with Curve Optimization.
there is a reason why GAME DATA SERVERS now replacing their intel CPUs and replacing them with AMD CPUs
and here you have the answer, if AMD CPUs have the same problems as Intel
I think there is something to the concerns with AMD, but this is mainly to do with the X3D chips. The Chiplets issues with the 5800X3D and default BIOS settings were an issue a year ago where they ran into some serious problems but it was only a mild discussion point any no one was actually panicking. I switched from the 5800X3D to 5900x and s=decided to watch and wait on the more recent offerings.
I would have to disagree with this AMD has stated they find the max settings of their CPUs and then purposely turn them down a little bit to prevent the type of problems intel is running into right now even if it will cost them not being number 1 in the benchmarks they believe their customers are better served this way, and after RMAing my 13900k last week i would agree
Havent done any real digging on my 7600x performance like this, but i do know that when i game, its rare for it to exceed 75C, let alone getting up past 90
I'm glad I haven't upgraded my 12900k, I have a z790 and was thinking about dropping in a 13900 or even jumping to the 14900 last month but saw all the issues popping up and decided to hold off until more info came out. So far my 12900k is stable OC'd and nothing weird happening.
Just tead the question posted for the video, " will all intel and amd cpus degrade over time?" My answer is question based. Will a mechanical engine fail overtime? Do spark plug wires degrade over time? If you answered yes, then you understand, nothing last forever.
For a long time I have felt the same way about where products are going...really I don't want to see temperatures above 85 degrees Celsius whether that be CPUs, or GPUs (7900xtx cough cough)...while most of the time the 95c temps come with short term heavy workloads like cinebench which really are not meant to be sustained for longer periods of time...I am in 100% agreement that everyone needs to dial things back to the long term safe zone. To put it in perspective PCI-E 5.0 M.2 NVMe storage drives will throttle starts around 81c...that's with a huge night and day difference in total drive power draw. DDR5 has a max operating temperature of 95c...so just imagine this theoretical scenario... the CPU next door is at 95c, the M.2 is at 81C...and the memory on the gpu is hitting 110c...bad stuff will happen to those DDR5 memory modules running at 1.4 volts with tight EXPO or XMP timings/subtimings. All components are starting to run hotter...and that confluence could very well lead to premature degradation, again they should just set the limit line at 85C for CPUs and GPUs.
You're bullshitting about AMD. I think INTEL'S manufacturing is the problem.
aside of always use intel i also always use radeon
both companies let two cores boost beyond what the silicon can handle over time, they most likely do it because we consumers are dumb enough to like the higher numbers on the charts of techtubers when they do game tests without considering that those numbers are achieved by detrimental clocks. Most that have locked the cores to disable the boosting from day 1 have zero issues with their cpu's. The sad thing is that if one finds out too late then the damage could already have been done and there is no saving the cpu.
I bought a used 5900X some days ago and I set the PPT to 105W (from 140W), the max temp limit to 85°C (from 90°C) and the curve optimizer to -30 all cores. In the end I lost a couple hundreds points in cinebench multicore, with the turbo still pushing to 4.950 GHz.
Waiting for AM6 for the upgrade.
@@gianlucapx the used price of the 5900X makes my 175$ purchase for a new 5700X early this year look silly 😂. Got it running just under the speed of the 5800X with a +200mhz offset and PBO curve optimizer voltage offsets of between -10 and -25 on a per core basis. 😂 Max wattage set to 100 in bios and it's super stable and runs very cool under full load with a peerless assassin cooler. Waiting for Ryzen 12000 to upgrade 😂
Set it and forget it! Forget overclocking!
Well.. as Wendell pointed out this degradation also happens on stock settings so setting and forgetting ain't a good idea. Check recent Framechasers video. You need to lock cpu boost clocks to something resonable.
This sounds like a clickbait for AMD CPU users. Because if you've ever used one, you know that AMD has placed hard thermal limits on their CPUs. I wouldn't call that "pushing it too far".
First the burning power cable of 4090 i guess. Then this intel oxidation problem. 💀 Amd is not really far away.
🥹 What the hell happening to the tech world.
thats why a limit my voltage in the bios, and PBO takes care of the rest, upon trying to get highest score on benchmarks and core clock when I first got my R9 5900x, I saw it reach 209w and just below 90 C, but I did not let the test run for long, I tuned way down... I get around 4.7 GHz in gaming and I never seen my processor pass 72 C and that was a very hot day in my room... so I'll be upgrading to the new 9000 series probably, and I will do the same thing, limit the voltage and let PBO do the rest, if I lose 2, 3 or 400 MHz on boost clock so be it... this matters for higher level consumers, devs etc... as they push these processors a lot harder than most of us will ever do
Just intel so far. Haven't heard anything about AMD having these issues of failure.
How do you lock all cores and prevent boosting on the MSI BIOS?
You're just making up controversy about AMD based on nothing. Per wendall of Level1tech , AMD accounts for 0.31% of cpu hardware errors.
Laptop chips ROUTINELY surpass 100c and it's been that way for quite a long time. Intel's chips have a DESIGN FLAW, exacerbated by the fact they continue to cram ludicrous amounts of cores into an undersized package size while routinely allowing the cpu to pull 320-450w. I'm old enough to remember when pentium 4 was the big thing, and we happened to have a compaq laptop that had a desktop variant of the 3.2 ghz pentium 4. Which was rated at 82w. For 1 core.
The barrel jack on the laptop got hot enough that the solder joints connecting it to the motherboard....MELTED CLEAN OFF!
Not once.
not twice
But THREE TIMES.
Despite limiting the power profile.
The solution was to take the power connecter from an old Compaq scanner keyboard and solder that to the board to put some distance between the solder joints and barrel jack.
The only other Intel chip i've owned over the past 30 years, was a core i7 940 on an x58 sabertooth. Never overclocked the thing, yet that thing managed to pull enough power that the VRMs simply crapped out. This isn't a new problem for intel, it's just the most egregious example.
AMD cores are pulling between 10.7 -15.8w each under full load, depending on power limits, intel's P cores are pulling 25-35w each.
There's no comparison to make here, particularly when every generation of Ryzen chip for the past SEVEN years has had a max temp of 95c (save for the x3D variants that drop the temp to 89c) where as the 13900k launched less than two years ago, the 14900k has been around for TEN MONTHS. The only issue AMD had was due to motherboard manufactures failing to properly limit power profiles, while intel has encouraged board partners to push everything to the max out of the box up until the point their chips began failing in droves....and yet even with more restrained power profiles newly purchased intel chips continue to crap out.
The problem is: the CPU are NOT getting ENOUGH voltages!
Motherboard Manufacturers Power AC and DC load-line settings in the Bios are changing the voltages that the CPU gets,
thus not providing enough voltages at certain frequencies under high loads.
I can run the 12 gen CPU on Load line calibration LLC 1 fine, but the 14 gen needs LLC 8 to achieve higher frequencies because there is too much V droop at lower Load-line settings, and the CPU does not get enough voltages under higher loads..
14 gen needs more power and voltages under high loads..
So Motherboard manufacturers just needs to fix their Bios settings ...
And it is cool that intel lets us change the load-line settings so that we can under-volt the CPU, but if the under-volt is too aggressive crashes, errors, or instability happens.
i tested this, and even the "out of video Memory" error was not present when i just was sending a bit more voltage to the CPU, but it was present if i was sending to little voltage.
Apps and games were not starting or crashing with too little voltages, but not crashing anymore with a bit more..
Yeah right don't be blaming AMD for intels bad choices cutting corners. AMD cpus are fine and don't have failing issues like intel.
Does this affect every intel chip? Is my humble i5 13420h in my laptop gonna bust too? That would be lame.
I can almost gaurenttee that all cpus build disposable just like all things corperations produce. is part of their design. they do not want products to last too long
Everything degrades overtime and nothing last forever 😂 just controllable or as planned or not
Never this fast. I saw in January of this year 13900k die when my friend bought it at a launch.
I just had a R9 5950X die on me suddenly one day... I've been gaming on computers since 1995 and this is my first and only cpu to die on me. The thing that is odd is that this cpu I have that died lived a very pampered easy life... never overclocked and has been liquid cooled so never hot either. Rough luck I guess....
7800x3d, PBO under/overclocked watercooled system. Had it since Jan. Run it 8+ hours a day. 0 performance degradation or issues. Also 6200 ram on the mem bus. AMD solid so far
I think X3D is at the lowest risk of all. They have much lower voltage and clocks.
7800x3d is not that old. Let's talk again over 5 re-pastes.
@@Giovanni-Giorgio The intel chip failures come faster than the length of use ive ran mine at. Also Im using a cryosheet, so there will be no repastes.
Sorry, but you are making shit up, amd's returns rate is minuscule. Why are you making crap up???
I had no idea about the Intel 13th and 14th cpu problems until my new prebuilt with a i7 14700KF I bought last week started to BSOD several times.. Thank god it had a 30 day return policy.. I'm most likely will never buy an Intel processor ever again. AMD's new processer releasing next month will be a big hit!
Intel CPUs degrade because of HEAT. This is mostly because Intel and motherboard manufacturers pushed voltages and power consumption of CPUs too high for die to transfer and dissipate.
AMD doesn't do that. In fact, heat production (that TDP number) out of the box in AMD CPUs is same as it always has been since Athlon X2 (early 2010s) days. This means, AMD's CPUs don't degrade nearly as much and are usable to end of their usable life.
I still use those old processors to my daily computing needs and those still work just fine.
The high voltages and click speed theory has been debunked by Lendel at Level 1 Tech.
Server class boards running 13 and 14 gen CPUs also have problems and they are built for stability
It ain't up to engineers. I blame stockholders,roi,and capital gains.
Didn't you hear what Intel said, there's a problem with fabrication process, something about base for plasma ion coating.
I use my AMD and Intel cpus according to specs, for plus 25 years.
No cpu ever died, but everything else
Nobody knows why Intel CPUs are degrading so quickly. People talk about the over-voltages on the motherboards (which was real for many of these units, including my 12900K), but there are units which have NEVER been overvolted or overclocked, which degrade just as fast. It is 100% speculation at this point, and it is highly improbable that your local internet celebrity / yahoo will have the knowledge, equipment and experience to figure out the actual reason. Hopefully Intel engineers are burning the midnight oil on this, as this seems to be shaping up to be a huge problem.
You know how old school CPUs could last you 20 years if you wanted to? Intel and AMD realized that most people are upgrading every 2-5 years and will never have it that long. So, why not push them harder, knowing they’ll degrade faster? Who cares, they’ll be buying a new one soon anyways.
The problem is, I don’t think they expected the degradation to be this bad, this soon. They got caught with their pants down.
I think they knew it would degrade faster, but miscalculated the rate of degradation
What the man says makes sense, and to mention it again, no AMD doesn't seem to be having the same issue as Intel.
That doesn't mean they won't ever get the same problem if they continue this how shall I say ummm, pissing match, and it is that, when the makers knowing set limits to high, they willfully are the bad guy, and for what?
The title of we're the fastest? Who cares when the CPU is fried, it goes from the fastest to the deadest 😂 Perhaps AMD will learn from this mess, maybe they won't. Let me ask this , why pay a good amount of money for a CPU that's going to kill itself off much sooner than it needs to? Don't be sheep people you don't have to buy the biggest, or fastest anything to achieve a great computer experience.
As long as you modified the the AMD cpu's with manual eco mode settings you will be fine..
The more interesting mystery is why do idiots still buy intel nowadays.
From level1tech data Intels problems look to be memory corruption which could be cache or Memory controller hence down clocking and dropping to single data rate not double data rate is helping as strain on Memory is reduced and Memory controller has more time to refresh memory cells. I'll speculate It could be Memory controller can't charge Memory up to level for 1 to be maintained and it's dropping to point it's seen as a 0 and data is then corrupted CPU slows as it trys to correct that then crashes which is symptoms Wandel of level1tech was talking about so no software would fix it unless it's to slow cpu down so controller has time to manage Memory.
It is an intel exclusive "feature".
My 7800x3d never really goes above 1.1 v
Just run AMD in eco mode. Problem solved. This is the mode they should be using by default. With Intel? Do you even have an eco mode? I don't know. My pc has never crashed. In 14 years. I push nothing. But I will need to upgrade at some point.
We are the quality control department
beta tester
kind of want to put it in a case, for peace of mind
i never would let my cpu go to 5.7 ghz lol, i like below 70 degrees . thats why i just use 7800x3d
Doubt it will happen to amd. Intel's always get hot and I feel that heat is the primary issue
Total bullcrap. The cpus crashing are at least 97.5% Intel
Anyone thinking architecture shrinking brings only positives but what are they thinking? If something is small enough it can be killed with really low voltage and amps if architectures continues to shrink this will happen more often and sooner after buying cpus gpus or whatever so i think architectures started to hitting physical limits even if they are capable to operate at lower voltages and amps and every chip will fail someday that is for sure and anyone did not mind radiation in nature that can also affect lifespan if tranzistors are small enough and exposed to it for a long period of time 20A will be not really affected or measurably at least i think but what about nanosheets or whatever...and i don't talk about pcs at home but in hospitals where are some radioactive isotopes stored and used will they develop some form of shield for pcs?!
You know, I'm still playing video games on my other old pc.
it uses an AMD Carrizo CPU/GPU, I think A8-7410?
Which ones have issues? Only i9's? Are i5 and i7 safe?
i5 are totally safe, i7s have some bugs but it's not frequent.
@@Hatsunari_Kamado ok, thanks
@@Hatsunari_Kamado @_mikkoxd_292 this but still, like Wendel said at The Full Nerd Podcast, update your mobo's firmware even if you have a 12th gen Intel CPU.
lol if your CPUs have degraded that means that you ran them stock or let the Mobo auto OC and I thought you're an enthusiast.
I've been running my 9900k OCed to 5.1Ghz all core and 4.7Ghz cash at 1.31V with 4000 cl16 RAM for over 4 years and no degradation at all, I'm very happy with the performance and have no plans to upgrade anytime soon but that said don't think it will be to an Intel CPU, probably going Rizen x3d next, especially if it can be OCed.
Learn to tune your hardware, it takes 1 time a few hours at most and you're set for years.
I'm waiting next year for a price drop on those new AMD Ryzen 9000 series.
Also i don't buy anything without watching reviews.
One reason I prefer manual OC with manual voltage
We all know that all CPU''s degrade over time its the reason why my 3770K can't even run on default clocks without increasing the voltage after it was running on OC for around decade the issue with Intel right now is that the processors are degrading fast and its on level never seen before AMD doesn't have that issue they have been pushing it for a while now and have been cranking those voltage since Zen2 I have seen multiple motherboards crank up to 1.5v even when idle and those chips are still alive and well to this day I am not saying its not degrading them cause I have had a 3600X that won't work on default clocks without cranking the voltage to 1.4V and it wasn't even overclocked if your someone really worried about degradation then you need to look into under-volting and using the curve optimizer brings more performance with lower temps to boot.
Celeron and Pentium won’t degrade 😂
Ill tell you what i think, but first a word from my sponsors for the next 10 minutes... thank you
why are they using a ring bus instead of a mesh?
Take a 20-core chip like the i7-14700. With a ring bus, you need 1 set of transceivers per core, for a total of 20. With a mesh, you need (19*20) 380 sets of transceivers AND each core needs extra arbitration logic and buffers for the 19-channel multiplexers. Not to mention that they can etch the ring interconnect on a single layer of silicon, whereas the mesh will need many more layers than that just to route it. The mesh will be somewhat faster during high utilization, but even at peak loads most of the bandwidth will be wasted as its highly unlikely that every core needs to send data to every other core. They probably did a cost-benefit analysis and the answer to using a mesh was probably "OH HELL NO" after only a few minutes of thinking about it.
I expect zen5 will be such a banger of a cpu and playform. Lower default power, better pbo and boosting, amd clearly sandbagging to throw off intel(although.. if i can see it, intel can see it). I expect +20% in games and then a +20% for the vcache version on top of that because they cut the latency of the cache system drastically ao we will benefit even more from the stacked cache. Yall get ready for an absolute bloodbath. Intel will be out here selling "similar performance" cpus for half the price of amd because no one wants them. Heck.. their margins in current cpus are horrible compared to amd so they wont be able to profit at all. Im sure they will come back at some point but just like amd after bulldozer.. Intel will have to rebuild trust and find ways to build more efficient and profitable cpus.
100% failure rate of 1x cpu, very representative
When i was in computer school
I saw size of chips schematics
If drawn on paper it was size of huge wharehouse
They shoukd shave bad cpus see if there hot spot
My thoughts there line too close long or bent to sharply
2 nanometer you get em interference so it needs em shielding
This prevents hacking and malware
My motherboard acted up and toasted my hhd
It is also good that Intel wants to hold unto their semiconductor factories thru placing 20 billion toward their advancements to near atomic sizes.
Where node size competition will soon be irrelevant when near atomic node size have been reliably reached in the near future.
This also means that once this happens. It will not only allow Intel to produce equivalent quality products at superior prices than their competition can even afford to do for them having to rely upon other companies like TSMC to produce their chips. Thus forcing competition pricing down to a minimum in a world that cries for economic prices.
But it will also mean that stability issues will be less likely thru that as well. Especially for more performance with less heat and electricity needed too.
Thus Intel's success and holding unto its semiconductor factories will certainly benefit seemingly everyone, save but the Nvidia guy and his hopes to sell everyone graphics cards at nearly 2 to 4 times their value of course, hahaha.
Hmmm, do you member 5Volts 8088's? 80186 (mostly used as MCU not CPU, HDD controllers in particular used this baby many a time) , 80286, 80386 and the big majority of 80486, only with the Pentium family did we start to see lower than 5V core volts as the new normal, specifically 3.3V, we have to go to the Pentium 2/3 to see voltages at 1.x volts.
So, you are speculating like there's no tomorrow, that said the CPUs are ageing way too fast if this is not a factory fault (well would still be factory fault, just a different one) it may be a raw metals quality problem to start with, which if true means only the factories receiving lower quality metals would be impacted which would completely invalidate your AMD theory, and yes theory because there are 0 confirmed facts here for it to be anything else.
fint think so bevause amd are from tsmc but intel crrestes their own, intel may have made s mistske somewhere in the
ir production process
New amd cpus are 7 degrees less than the 7000 series.
All Microchips degrade.... Intel just the fastest. Few years ago Nvidia when they supplied GPU's to Apple 😅
I am a computer builder i build on ryzen and intel platforms let me remind everybody ryzen 3000 cpus where also dying at a high rate because of pbo pushing over 1.4 volts through the cpus. Nobody remembers this because its amd. Lock down your cores never exceed 1.4 volts and u will be a happy gamer for a long time no matter the cpu
You're wrong. But that's okay, Intel's 10/11 and even 12th gen are fine, the reason why they degrade is down to poor engineering, they moved the I/O layer, which us why this only impacts 13 and 14th gen. Don't forget that click wise there's not much if a difference between a 12900K and a 13900T fir example. Bad engineering is what lead to this. Nothing more
Hey is the i5 13600k also affected?
i wonder about the 3d series processors since they run a bit hot on average but everyone I know with one including myself either turns off all "performance boosts" their mother board has and runs them bone stock or under clocks them.
That's why they are clocked lower than non-3d counterparts and conventional OC is disabled.
This has been happening since the 10th gen.
11th
Every thing has its limits even our planet where we are livin in. Degradate is what we are..slow or fast is a another topics for einstain theories.
Are any old model CPUs affected?
!2th gen are fine not showing widespread issues. The 13th and 14th gen I5 and I3 are fine some I7's have reported issues but that isnt widespread
Not enough data to know yet. They are likely at far lower risk.
ryzen 3000, funny how he mentioned it but wasn't aware it had this problem years ago
@@asdasdasd-gx7zs i had to lock my voltages, because i use a 3900x. My 5950x was screwed up because of voltages and i had to get rid of it. Can't say anything bad about amd apparently.
Still running an i9 10900K at 1.5V via Asus’ AI OC since release.
Never been an issue.
@@DainHunter bro u running a custom loop? My 10900k only needs 1.38 volts for 5.1ghz all core to be extremely stable. Why the high voltage?
@@betawing1914 I am running a custom loop indeed. I just want my CPU to be as stable as possible.
What can I say? I thought competition was good.
All CPUs degrade over time, just that Intel is degrading exponentially faster...
If you would do your research instead of parroting you would find most of the fails are on Asus MB. Asus as default pushes 100A beyond the specs. Of course it will fail.
uh, hope wont take long
absolutely right i agree we are pushing cpu and motherboards to way fair hince the issues arising now we are trying to get somewere not really needed i think the issues are vendors in all sectors in terms of cpu motherboards and graphics cards needs to sort of work together to come up with a solution on all sides that work inharmony and untill that happends the same issues will arise and yes its easier said than done but if business means as much to them as they cliam then thats the only way to sort and fix also theres a big enough market out there to satisfie the market but really oh iam kinding that wont happen so problems will keep happening untill gives hay its happend many times and will keep going and going till it ends up blowing up in many faces literally and hypothetically thats when they will go whoops we should of gone an other way and by that time too late so let see what happens hopefully they find a way to sort it but cant see it 🤔🤫🤨😒
ahh i missed the snort.
Zen 4 was TESTED to run at TJMaxx at 90c
I think Ryzen chips will be good.
x86(x86_64) is long death. So engineers need to thinker about new tricks to still get some performance boost over the last gen, so they can still sell something. And both lie and lie and lie about it. You know that you are at the end of the road, when only more cores, more and faster cache comes to rescue. Would be funny to compare each generation benefit on the time an single instruction needs in ns. How many years was needed to go from 1Mhz to 1Ghz? How long from 2Ghz to 4Ghz? And now we fighting the 5Ghz Bullshit war. Amdahl's law kicks in, and apple keeps smiling with there non x86 cpu's.
I never oveclock my cpu and never had degrading issues😂also not using pbo
7950 x 3D 5.8ghz. all core, posted a video on this dilated, liquid metal.... Still 100% stable today. Please consider changing your name to mentally challenged. It would be more accurate.