So the thing is intel programs a voltage frequency curve into the CPU. When the motherboard removes the power and current limits. The CPU will request insane voltages because it's not hitting the power or current limits. So technically the motherboard isn't feeding more voltage than the CPU is requesting. However if the power limits were being properly enforced the CPU wouldn't be able to request insane voltages.
@@georgejones5019Yes and no. Intel is different in how it works because they have a set power draw curve. Specifically what ASUS is doing here is tricking the CPU into a false understanding of it's own state to get around Intel's hard coded limits. The CPU requests more voltage because it misunderstands what the MB is doing. In some respects it's worse than the AMD problem. AMD will mostly just let you send it whatever you want and it's up to the MB manufacturer to stay within the spec AMD gives them. With Intel they are purposely bypassing and taking advantage of a sanity check Intel put on the chip.
but you're not describing "optimized" you're describing "fail-safe". 100% safe and reliable with no overlocking should be the default, but not the optimized default.
@@brenthauer8365Yet, there's a huge difference in between 'unstable' and 'destroys your hardware' which 'optimized' does not account for, neither in tradtition nor in meaning, because it has been there in the BIOS for decades and never killed your hardware before, and the word 'optimized' doesn't include 'destructive'. 'Optimized' never even came with overclocking in the traditional sense, it at best decreased RAM timings and set some little optons like AHCI instead of IDE and disabled energy saving features and whatever - but nothing even remotely endangering the integrity if your hardware. And the most important thing about this is to acknowledge, that the vendors are doing this intentionally. They want their boards to be just the little bit faster than the others' boards in reviews, thus making unsafe settings the default. It would be very easy to simply add other settings like 'Dangerous' or 'Overclocked Defaults' to any modern UEFI setup, but they obviously just do not *WANT* to do that. They actually are misleading customers by lulling them into a false sense of security. Customers who pay many hundread or even a thousand bucks for a mainboard - for then being treated like idiots as a reward.
@@brenthauer8365 The problem is most, if not all, modern motherboards only give the option for "optimised" defaults. In this case being the only option they should be 100% safe and reliable without any auto overclocking regardless of whether they use the "optimised" nomenclature. But for argument's sake let's say a motherboard has both a "fail safe" and "optimised" option I would still argue that optimised should still be 100% safe and reliable with any automatic overclocking being kept well within the limits that both the CPU enforce and what the motherboard is capable of, it certainly should not be disabling any limits nor should it be trying to push those limits to the extreme.
@@brenthauer8365The point being made is that the optimized defaults IS the default setting out of the box, so these optimized settings are the ones you get when you buy your motherboard and first use it and also whenever you update your bios. The only way to not have these optimized settings is by physically changing bios settings. These should not be the default settings. That is the point of this video.
I had a 10th gen ROG board fresh out of the box on 10th gen brick it self during the initial setup because it was pushing “unstable” “”stock”” options. Ironically I was building 2 systems and the other would crash **every time** installing windows. Turned it off and fixed the issue. Sent them both back. Eff that noise. They haven’t learned anything.
They want to sell you a new CPU not too long into the future. They have learned that a lot of people still use a decade old CPU and it Works! ...due to low silicon degradation, and that they couldn't improve sheer IPC that much. So the new business solution is to degrade (fry) the silicon so it only last trough warranty.
Asus is the worst for this... I had a sandybridge 2500k that got slowly died as the bios kept trying to overclock it... it's their buggy AI Tuner crap. Which is still at it on the 7950X with AM5, I've had to set an artifical voltage ceiling of 1.3 volts.
To be honest, Asus is doing intel the favour here. U got more performance than what it should(intel power limit). Of course at a cost of power consumption. Which lead to other problem like fan noise, depending on ur cooling budget. If all intel unlocked processors got benched by limit enforced.. its multi-core performance will get affected by quite a margin. And will look less desirable. Already at tough competition with amd..
I'd like to add this to the average viewer who wouldn't know this information as simple as I can put it, but Voltage (Volts or V) does NOT generate heat when applied to devices (CPU, GPU, RAM etc), under no load conditions 1.1v and 1.5v will give you the same temperature assuming no load as long as it's within specifications. (Including windows background processing) However, current (Amps or A) are what causes heat generation when it flows and power is present (Watts or P = Volts x Amps) through the devices and potential damages can happen when current isn't limited or regulated using the voltage applied. So what those motherboards are doing are basically removing the TWO PROTECTION limits of current and power, which is... Yea, very harmful for electronics...
To clarify, no motherboard should set default limits _higher_ than the CPU manufacturer recommends. If you have an A320 board and want to drop a 3800x in it, the board should absolutely enforce its own lower limits even if that means gimping the CPU. Better that than cooking itself.
@@grgspunkI remember dealing with that before upgrading to an MSI board last year, I stabilized the ASUS setup by giving a somewhat negative power curve to the 7950X3D.
@TheDivisionAddictMy AM5 B650 Elite AX at $200 is much more solid than my X570 Elite AC was at $200. Gen 5 on one port. I guess it depends on your buying segment. I wouldn't say that is the market as a whole.
Oh my God... For the past 2 YEARS I have had issues with CPU temps under extremely minimal load. I have changed thermal paste, fans, AIOs and even bought a new CPU with no luck. After literally disabling one setting as suggested I am finally at 38c with 30% CPU usage. Thank you, thank you, THANK YOU
I'm going to assume its an asus board because its a 99.999999% chance that it is. Until they learn their lesson and it affects their bottom line they won't stop, it allows them to "win" benchmarks at the cost of your hardware. They know that the majority of "reputable" reviewers now only do default settings so they choose to do this on purpose.
I just recently built a gaming rig/everyday family truckster with an i7 14700K and couldn't understand what the deal was, went from a Cooler Master 212 Halo to a Corsair H100X AIO to try and wrangle the outrageous CPU temps while playing Helldivers 2 and Arma III. So I wasn't losing my mind, it really is a voltage/heat issue stemming from the MOBO. Gonna dive into BIOS when I get home from work and investigate. Keep it up Jay!
@@SinisterSkyler Because you can get the same performance with less voltage, its like the reverse of overclocking, you just keep pushing it down until it crashes, then you bump it up a little and let it run at that voltage. This can keep your temps down by roughly 10-20 C just with undervolting alone.
Jay is an awesome technical content creator. With him, I improved my English listening skills, learned how to maintain my desktop computer, adjusted the tweeks of the components, and understood the synergy between them. I heard that he is passing through a delicate medical situation. I wish him my best vibes to get over whatever condition he is in. Human being like him are what society needs to understand the meaning of "greater good" for being empathetic to others and helping others. Thank you Mate!!
thanks2, glad to be part of your success, still its hard to keep positive mentality with this weirdness, fact is nobody give any $ which nobody cares, and if they still argue w classic 'afraid this that' then they also dont care abt me
🎯 Key Takeaways for quick navigation: 00:00 *😓 CPU and motherboard default settings can cause overheating and throttling issues.* 00:41 *🧭 Motherboard manufacturers apply optimized defaults that overclock CPUs and increase power/voltage limits beyond Intel's specifications.* 01:23 *📝 The video aims to explain why some users experience high CPU temperatures despite adequate cooling.* 02:19 *🔍 The issue is specific to Intel CPUs, as AMD CPUs are not pushed as aggressively by motherboard defaults.* 03:30 *⚠️ Enabling "Let BIOS Optimize" sets higher power limits, amperage, and voltages than Intel's specifications.* 04:42 *📢 Motherboard manufacturers should load Intel's default limits out of the box instead of aggressive overclocking settings.* 05:53 *🔋 Motherboards provide more voltage than the CPU requests, leading to higher temperatures and power consumption.* 07:45 *💻 High idle voltages are normal for CPU stability, but load voltages and temperatures are concerning.* 10:22 *🔄 Resetting to Intel's power limits results in lower temperatures and voltages while maintaining performance.* 14:07 *🚨 If experiencing high temps despite adequate cooling, check and reset motherboard settings to Intel's defaults.* Made with HARPA AI
So general consensus is to deregulate, but the solution is to make them feel it in their wallet where they also get a bunch of money through B2B deals anyway…. Got it This might work with game developers but it doesn’t really work in the main space where these companies have diverse portfolios. Oh you’re not buying their motherboards? Just focus more on servers or something.
@@methane1027 20 years from now, it will be boutique in the west to buy a custom built salvage motherboard. The only thing stopping that will be anti-repair laws
Just built a 14900k. Posted Aida64 stress tests to "it humor and memes" and had peeps concerned abour my 100c spikes. After your vid, went back to the BIOS and sure than shit had the defaults on. Enabled the Enforce limits and voila didn't go over 87 on a silent fan profile and 82 on max fan. Posted the results to Twitter and gave yiu a tag. Thanks a million for this vid.
I bought a Alienware r16 with the 4090 and 14900, it’s running hot too on demanding games it constantly is at 87-92. Smh I’m new to pc gaming. Scared to mess with settings like this
@@donjuan8124 87-92 on a PreBuilt I think is pretty good. Not sure what kind of overhead you have with the Dell Bios, but Prebuilts are know for spectacular air flow. I avg 87ish now OC'd to 6.2. Fan settings are on quite profile as well. So I'm happy with it.
This is amazing. Just built a new system with a 14900k and was baffled by the issue you described exactly. I'm not experienced with overclocking and voltages etc so I was a bit stumped as to why I was maxing out my temps then as soon as I took off the load dropped to like 30 degrees. Thanks so much for the vid, the timing was superb!
I don't leave comments to often, but WOW. I have been having this issue with my i9-14900k and had no idea how to fix it. Repasted and replaced my cooler multiple times. This simple setting on my ASUS MB solved it right away. Max temps now are around 85C during a Cinebench Multi Core test. !!!THANK YOU!!!
Did you change any other settings or only the settings in this video because i have the same specs as you and i followed the steps in this video but still getting 100c. Replaced the cooler and still same issue
THIS!!! This was me! I spent MONIES on making my first custom water loop, and I thought i had done it completely wrong, because every cinabench run i was getting thermal throttled... I could not figure out what i did wrong so tweak i did... and Boom its under control... pulling 350W at max settings on my 13900K at 96C 6GHZ all cores... so yeah my water loop works... it was the no power limit causing the issue. Could not have fixed it without Jayz help. Thank you Buddy!
I need your help. After disabling Asus MultiCore Enhancements, I only get like 5200 or 5300mhz on p cores when running cinebench. Also the current cpu core/cache current limit is 500 Ampere. According to Jason it should be 360 amps. What am I missing here?
You need a bigger or second radiator and/or a better water block. My AMD 3900x water-cooled stays in 50s to mid 60s C peak under full load. I have a custom loop with a good water block but an insane radiator - the Alphacool Nexxxos Monsta 560mm. My case actually supports two of them but I only use one for my CPU and 3080Ti. That radiator is rated for over 900 watts of cooling. I am currently planning to upgrade to AMDs Ryzen 9950X or 9950X3D when they come out and at least know that my water cooling loop which I designed and built with my 3900X upgrade will be sufficient for many more years to come. I've always believed in overbuilding on important aspects like cooling and power capacity. It makes for a much more stable experience.
@@Fendera1you could start by asking the question in an intelligible manner. No one can see what CPU you have, what cooler, what motherboard, etc. no one will take you seriously if you can’t be bothered to provide that info.
holy crap, this is literally a breakthrough video for me. I have been struggling with crazy temps for a while, and even went through all of the trouble to figure out how to UNDERvolt a i7-12700, but i didn't realize that hitting DEFAULT was actually over powering my processor.... I thought I was going to base settings by pressing defaults, but obviously not technical enough to read through the menus to figure this out. I have just found this video after changing my thermal paste, moving around my fans, considering replacing a 240mm Arctic Cooler Master II, this was the fix. Thank you!!
The only 2 things that will cause companies to stick to default limits are 1) Everyone refusing to buy boards that ignore default limits, or (more realistically) 2) Intel telling them to knock it off or they don't get to make LGA whatever boards anymore.
Look up ECO mode and how to set up PBO values for your CPU if you're using and X-version Ryzen. ECO mode will reduce your temps at the expense of speed, but if you're just gaming it's not noticeable. Personally, i'm okay with an avg of 1-2 fps loss to go down to 83 degrees under load vs 90 by default, but most of the time the fps loss is not even there either, so it's jst a win-win. That one's not on the mobo manufacturers, but on AMD. If you want to be extra thorough, you can try undervolting along with it, but ECO mode itself will do just fine. If you're not on an X-ver Ryzen, you don't need to do anything, but you can always try undervolting.
There's not such an issue on AMD CPUs (AM4 at least) you can play with the curve optimizer and PBO to get quite good gains in both performance and temperatures.
Thank you for this helpful video!!! I just built a new PC last week with the Asus Rog Strix Z790-e Wifi II and used the Asus Load Optimized Defaults. When I ran Cinebench my CPU was running at 99C!!! I then changed Asus MultiCore Enhancement = Disabled- Inforce All Limits and now the max CPU Temp is 72C. Thank for keeping my new build from frying itself!!!
Bro!! I made that one change and when gaming… went from 80.90C to 60sC. OMG!!! I was wondering if it was my water pump, etc. definitely saved my CPU’s life….
@@jeezusjr same thing happened to me. Bought a new Mobo, CPU, psu, cooler, and ddr5 ram to upgrade an old system. Hopefully get some of the $ back from returns Lol
I also find it insane that because of these settings I have had multiple customers come in with Asus motherboards and a 12900K or 14900K and the Cpu was unstable and they couldn't figure out why. It would crash with strange errors, one of them even claiming it was a memory error, when the only issue was that the CPU couldn't handle the power that was being forced into it and so it was crashing.
i went out to shop for 14900K with z790 last week, and i met another customer who made the same build 4 days ago. he came back complaining his i9 going 99-100 deg in 5 seconds, and wont come down even at low loads. according to him, his super expensive AIO was at fault, so he just asked the store manage to replace his cooler, lol. I told him, its not the case, just undervolt the cpu and adjust till you achieve max freq and low temps. conclusion: he still went home with a different cooler 😅🤣
@@deepak_nigwal same thing happened to me i got a z790v board with a 12900k i got a water aio cooler. Took it to microcenter and they installed my cooler wrong they.i was pissed but they fixed it and it worked but i be download msi afterburner everything fine but my temp says 120° cpu idol .i turned off my pc checked in BIOS and my temp was a cool 50° . But on msi days it was cooking steaks but my boss said my cpu is fine ????
@@deepak_nigwalthe mistake they did was they didn't adjust the aio coolee bracket so I'm thinking that was the problem bcuz I'll play cyberpunk and my pc runs fine for hours and after i check my bios ... Under load was 60° ....
In some cases I just had to turn off MultiCore Enhancement. In others I actually had to underclock the CPU by a 200 to 300 MHz and then it became completely stable. I had a computer in the shop last week that was throwing memory errors when launching games and there was no issue with the memory at all. The only issue is that the CPU couldn't run stable at the speeds it was binned for in certain circumstances. So dropping the core by 300MHz, basically running it as a non-K CPU was all that was needed to make it 100% stable. Before it was around 80% stable. @@byFraze
Thanks for making this video Jay. My 14900k was hitting 100c on multiple cores whilst gaming and benchmarking on "default" settings. Even though having a high-end 360MM AIO.. Once I disabled that setting I'm getting a stable 81-85c and the performance loss is only 2.1%.. To think for all these months i've been torturing that CPU for an extra 2% performance...
i set it to -30 on my 5800x3d, insane how much the temps dropped after that. still need to wait and see if it's stable but so far so good. it also performs even better now.
This happened with my 10700 on an MSI board. For the longest time I thought I had something thermally wrong. Chased it for a while with no results, then one day read about how motherboard defaults are ridiculous and, just like in this video, the motherboard defaults were nowhere NEAR the normal Intel limits for the CPU. Switched to Intel's base recommended defaults for the CPU and it instantly solved the thermal issues. Even aside from thermals, not only does it confuse consumers and waste time/money chasing problems not caused by the consumer- it wastes a ton of extra energy too - so thermals and energy skyrocket with extremely diminishing returns. It's so frustrating that this has become standard practice.
This but 10700k and two ASUS/ROG boards. One crashed during UEFI update time because unsafe thermal/power limits. The other crashed every time installing windows. The CPUs overclocked just fine tho. Returned both boards and went with a different brand.
Hi @neonvoid666 , actually just got through building a new pc with an MSI MPG Z790 Edge Wifi. So how would I go about setting the mother board to these settings you did. 13700k, 4070ti, LS720 SE Digital aio cooler. I would really like to do the Intel defaults. Cheers.
This is excellent information! When the max power limit was set to 4096 Watts (by MSI) instead of 154 Watts for my 13th gen CPU, gaming was extremely difficult with stutters and input delays of atleast 500ms, even walking/running in game felt difficult, counter-strafes never registered. After changing it to Intel's default 154 Watts, gaming is not painful anymore. Please make more videos on BIOS settings - Disabling C-States, Intel Speed Shift, Speed Step, Secure Boot.
Just built a new i9 14900kf system, with an Asus motherboard, 360 AIO, was having terrible temps, changing this setting really help. Really appreciate your videos.
If you're using a MSI motherboard, the power limits are auto set depending on what cooler you selected in the bios. If you select water cooled, it sets the max watt to 4095 like shown in the video. Change it to box fan even if you use an aio if you're having temp issues.
@@HuCuRuS hocam kapatmadan da boşta bile 100 derece zaten öyle olunca da performans düşüyor artık sıvı soğutmada var bir problem diye düşünmeye başladım
Great vids, always a joy to watch. I'm sure plenty benefited fully from it I also did but to an extent because I own an MSI laptop and their Bios is not exactly a walk in the park it's so so confusing, I really wish you guys will get the chance to expand it into the laptop segment too in terms of technical stuff like this vid where someone like me and I'm sure they're plenty because I looked and couldn't find anything in that regard when it comes to tuning an MSI laptop with an unlocked CPU variant through the Bios.
You are a life saver, I used ThrottleStop to address the motherboard settings issue, it was not enough, I still had power and temp fluctuations, even with base power PROCHOT lit up, entered the FiVR window, and set everywhere I saw 511.75 A, 307 A and the fluctuations disappeared. However, if I want to change the power limit from 125W to 253W, it doesn't apply it in ThrottleStop, Prime95 remains with 125W power, I entered in XTU and asked to reboot, before I did not need to reboot, I have to learn more about ThrottleStop looks like. I got a B760 mono because the Z790 one was 70% more expensive in Romania.
Thank you for releasing this video. I bought an R9 7950x like 8 months ago with a 360 aio... and it idles at like 60c... and doing anything that takes cpu skyrockets to 100c. Tech yes city mentioned undervolting because of how insane they are cranking the settings to compete. I dropped my voltage from like 1.2-1.4 (don't remember stock voltage) to 1.000 and it dropped to 35ish c idle the second I click the apply button. I have an msi motherboard. This just confirms I need to adjust my bios. I did absolutely nothing in my bios except maybe cranking my ram up, but I don't remember. Time to learn bios and fix it....
Pretty spot on and important. Went on a MSI Z690 from a non K 12600 on a decent aircooler to a 13600k and loaded "defaults" and at a sudden temps skyrocketed under stress. Couldn't for my bare @ss figure out what was going on but knew the new 13600k had some higher TDP so went and installed a decent 240 AIO instead. Well, the temps was still very high but not throttling but the TDP still "out of league". Then I started catch the info about "undervoltage your rig" etc and one of the most important settings on MSI was the "Lite load" settings that currently were at "12" I finely read the voltages and saw they were way too high! Played with the "Lite Load" and ended up with set it to "4" istead. WOW what a difference! Now I finally have the normal TDP 0f 180'ish watts and normal temps and still doing speeds as on the box and still rock stable! Shame on yours motherboards manufacturers litterally cooking the CPUs for marginal gains!
The latest MSI bios is a lot better about this, the BIOS my Z790 came with was pulling 220w on Cinebench23 with a 13600K but with the latest it's around 180w. I used the CPU lite load setting to lower the voltage a bit and now it's 150w and all cores can run in full turbo full time. I can get it down to 120w but I bumped the CPU lite load setting it back up by two increments to be sure of stability.
This video saved my life !!!. I just got 13700F and when i play BFV my cpu was hiting 100 degrees. I even decided to buy some AIO and other stuffs but when i disabled the ASUS Performence Enhacment 3.0 my max temp is lowered to 70 degrees. Thank you so much!
literally just went through this. Ran Cinebench with my 13600k w Artic Freezer II 240mm, was hitting 100*C almost instantly. I bought a contact frame, installed it and had the same result. Did some googlin, and found that my MSI mobo has a setting called CPU lite load, which was set to auto by default. Changed it to setting 10 (from 12) and am now maxing out at 90*C in Cinebench with more consistent clocks, and score within margin of error of previous. I could probably tweak some more stuff to improve it even more but it doesn't really matter for the day to day stuff, and it does idle/run games cooler now with the contact frame.
JAY YOU ABSOLUTE GUN! I built a PC about 4 years ago and used auto for most settings, I would often hit 100C and actually overheat even though I used a liquid cooling AIO. I changed some settings in line with this video, and I'm now no longer going above 55C or experiencing fans maxing out RPM for load Great stuff!
Bro wich settings? Im currently experiencing this myself, cpu just goes to 100 celsius instantly while being IDLE, I have a watercooler aswell and I have this pc 4/5 years now and never any heating problems, what should I do?
This video just help me out so much. For almost a year I've been fighting the temps on my 15 13600k and I just thought it was cooler/fan. After doing some of the suggested changes mention in this video my temps are FINALLY normal. No more 99c cpu @ 20% usage. Thank you and Keep up the good work guys!
Its almost like everyone needs to take a hardware generation off and just work on efficiency and having things work properly. Imagine if today, we had the performance we have but our CPU's still never really went over 60c?
There will be no over nine thousand power level and no 1.21 jiggawhats. It just appears to be a 12-bit value (not uncommon in ADCs and PWMs) which takes values from 0 to 4095, the latter value effectively meaning that the motherboard shouldn't be bothered by what your cooler is capable of. The same goes for the current limit of 511A. It's 2^9-1. These capablities may be used by very special, very small systems with small VRMs and heatsinks (think embedded, industrial, battery powered, aerospace, military), but are of no concern on regular mobos with 8 or more phases of voltage converter which is actually obscenely huge. I run my system with both settings maxed out for more than two years and it still didn't blow out the magic smoke.
Excellent video. I think one of the reasons this is still happening in 2024 is because it is simply not being talked about enough. So many thanks to Jay for doing so. Why having your CPU constantly above limits and thermal throttling to protect itself as a standard default setting is beyond me. The point being made is that it is fine to have this setting available should you wish to use it, but it should never be the bios default setting. For those concerned with performance (gaming), I would encourage you to actually test this. You will probably be surprised by the results.
Years ago I had a Intel I7 7700k and a MSI Mpower platinum z270 motherboard. Running the bios factory settings gave me a idle temp of 80c. Found out that the core voltage was set to 1.4v out of the box. I dropped it down to like 1.2 and idle temps went to like 30c. No real performance changes as i dont over clock or play around to much with it, MSI released a Bios update that changed the default to 1.3 as clearly they identifed this problem. I still run it as 1.2 and have no issues. This has been issues for years.
I did the same. I have the i7 7700k and a gigabyte Motherboard. My idle temperature was about 80. I decreased my vcore voltage to 1.1 and solved the problem.
Thank you! This is such an important thing to get out there. I feel so bad for people building their dream rig and then feeling like THEY did something wrong because of high temps due to motherboard vendor default OC profiles.
High temps aren't a problem on current gen. CPUs are now designed to turbo until they have to reduce clock in order to stay under TJmax. This is a good thing, you get more out of your CPU doing so, and your cooler is more effective when your CPU is hot (deltaT and all that). The old school users who are used to trying to manually keep temps under a certain level are having a hard time adapting to this idea.
I'm using an MSI z790i but their latest beta bios states "that it has replaced their own system power settings and replaced them with Intel defaults but users can still optimize system performance with alternatives from MSI." At least MSI has taken a step in the right direction, I typically shy away from beta versions unless it has a feature that I need.
Gosh! I spent so much time back in 2016 trying to figure out what was wrong with my i7-7700 (65W CPU) going way above 90°C all the time... I even bought a new cooler thinking that my Noctua cooler wasn't doing its job. I damned Asus when I found out months later that it was MultiCore Enhancement's fault... I didn't know they were still doing that after all these years. They must stop. This should be optional, not default. That video was necessary.
If you have an MSI motherboard change Lite Load from 12 (default) to 9, which is the Intel recommended setting. It makes a big difference in temps and scores.
Mine is also 9 but with 12700k, in previous bios I should set it to Mode 1 to be like I was when I first bought the motherboard... Now my normal is 7. You should just take a fast CPU bench with cpuz and check your score. Lower mode after till you see less score and then up it by 1 (or maybe 2 to be sure that no speed penalty exist)
@@haies09 That’s interesting. Mine was 12 auto. I just did a search and found reports of some people saying their auto is 12 and some where they say their auto is 9. I guess just always check to make sure.
To further this, go to Global Core SVID Voltage. Offset Mode Sign to Negative , Additional Turbo Mode to Auto and Offset Voltage to 0.10000. It will help to properly under volt your processor without hitting performance.
But if they don’t artificially inflate their “defaults” as faster than the competition, how can they claim their products are “better”? And if they don’t reduce the lifespan of their products, how would they make more sells if a system takes 3-5 times longer to fail?
@JayzTwoCents thank you so much for posting this video. I've been having horrible crashes, including some blue screens, with my newly built homebrew system. It's been having random crashes when under heavy load (usually cyberpunk or bg3), and hardware diagnostics couldn't find anything wrong. I disabled this setting, and I now haven't had a crash in two days. You are a lifesaver!
years ago and i am talking 25+ years ago they used to have load default settings and load optimized default settings. and the load default settings was the factory settings for that cpu. and the load optimized default settings was over clocked settings for that cpu based on motherboard manufacture settings for that cpu. they should be still doing this and i do not know why they stopped doing load default settings
Exactly my problem ..... I got Asus maximus hero x MB with delided 8700k on custom water loop and its hitting 100c and I was thinking i got bad cpu ..... started to look for replacement cpu and stumbled today on this video. Soon as I get home from work I will test the setting.
The crazy power limit is something I noticed immediately and quickly set the watt limit to 253. My 13900k stays at 67°C on cinebench r23. I also found gaming was noticably better with the 253 watt limit(as in movement/feel). Running on a MSI Tomahawk z790.
I loaded my BIOS back to default and did not fix the issue, but I was able to find a fix within Windows in power management and it resolved the issue with the CPU.
@@Chris256 You will need to go into control panel, then select change plan options. There's an option for change advance power setting, from there you will select (Processor power Management) If Maximum processor state is at 100 switch to 99. Issue will be resolved.
@@pjm7482 It will turn off the turbo boost doing that. Its a shame that we even have to do it bc some motherboards don't even let u change anything. And we missing out on performance bc we can't have turbo boost on without cooking up or cpu. I guess u could make your own powerplan to somewhat get out a little more performance without it hitting over 1.4 volt .
i don't understand, are the showed tips, e.g. "Gigabyte: Tweaker--> cpi upgrade: Default" your recommended settings or the settings that cook our parts? if they are the bad ones, which one should we chose?
I don't know how to pin a comment, but this one needs to be seen by Jay. speculate the video poster is the one who has that control. Confusing as he goes so fast.
Update: i downloaded cinebench and checked the temperatures in HWMonitor. The settings in Jays Video (e.g. for Gigabyte: Default & Auto/ enabled) instantely brought up the temperature to 100° C when i staeted cinebench. It didn't even take 10 seconds. So i guess the settings shown in the video are the bad ones, but i have no idea what to change, since in the gigabyte bios, the only other option apart from Cpu Upgrade: Default is "gamer mode" or "turbo" - which i honestly don't want to try out, since my flat is not flame proof.
pretty sure you need to set turbo power limits to Intel POR or else Enabled + your own limits The gigabyte manual says Auto is for intel defaults, but i think its incorrect and is actually gigabyte defaults and they edited the video based on that manual
OMG You saved my PC. I could cry, I have had this issue for years and even gave up with my pc for 3 years because of this issue. You found the fix, thank you so much!
So I went and checked and sure enough this is how my 13700k on a Asus board was set. I just assumed the default wasn't a oc. Dropped my temps in a cine run almost 10c lol
@@timothygibney159It didnt drop the clocks at all it sat 5.3 the whole run at 87c. Im fine with it how it is but i always wondered why it ran as hot as it did and this seems to have fixed it. Before it would shoot right up to around 95 96 and sit there i dont remember the scores i got last time it was awhile ago when i ran one but the test run i did after the change was still a little over 29k.
@@Premier024 mine I have to to prevent thermal throttle. It only has 95sp rating. I can hit 38k on cinebench now after the undervolt since it would down clock previously
@@timothygibney159 That's crazy 38k on the 13700k. Tbh I'm perfectly happy with it just getting what's it's supposed to with better temps I'm GPU limited almost all of the time at 4k as it is but the less the fans have to ramp the better.
@@Premier024 oh it’s a 13900k 😂. Close to the 13700k but shoot most other 13900k even with an overvolt can do 42k and 6.1 p core on at least 2 cores. I have 57 on 2 and that is it . I assume a good 13700k could hit 35k on cinebench easily if you adjust the volts and get at least 1 core at 6 gjz or 2 pcores at 5.9. I only light game old mmos and use this for virtual machines as a more cores are better for my use
My spouse is using my previous rig, with an Asus ROG Maximus VII Gene, a 4790k and a 980 Ti (had a 1080 but it ded). The Asus BIOS wants to run its adaptive boost, but honestly it cooks the CPU. Yes, it has been re-seated and re-pasted. It's gotten so bad it will no longer post with XMP enabled, when it used to run fine. Turning adaptive boost and multicore enhancement off is the only way that rig is still safe for gaming. I know that's older than the start time you mentioned, but it feels so related.
by re-pasted, do you mean de-lidded and the die re-pasted? don't forget the 4790k has paste as the TIM, not solder, so by now it probably needs replacing with something extremely viscous like thermalright TFX (the paste that's so viscous you have to warm it up so it even spreads. make sure to spread and have the sides covered too), or a honeywell PTM7950 pad if it's thick enough (the phase-changing cooling "pad" that GPUs are using on-die now instead of paste. it's only beaten by liquid metal but is much safer, but still pretty expensive. usually 0.3mm thick) edit: misspelled viscous
@@Efreeti might be worth looking into then, just like re-pasting an old GPU that still has life in it. there's minimal risk with a de-lidder tool (as opposed to the vice or razor blade method which killed mine from digging into the substrate by accident), but whether it's worth it to you both money and time-wise, that's not my call to make
A friend of mine had stability issue and bsod with the 14900k. He solved it by putting the cpu tdp and core amps at the value reccomended on the intel's spreadsheet instead of the automatic mobo value.
Hey Jay, please consider making a video where you test multiple different thermal pastes! Noctua NT-H2, Thermal Grizzly Kryonaut and Kryonaut Extreme, Arctic MX-6 etc. Especially because new pastes come out every couple months, it would be interesting to see which ones are the best!
Always nice to see that my i7-8700k is still holding up well after 7 years because I adjusted these settings and delidded it as soon as I got it. I only had to reduce my 5ghz overclock to 4.8 recently to maintain stability.
I have this exact problem with a B660 and 13700K. I had to workaround to get 1.38v, whereas itself gave 1.61 (almost melting socket voltages) and after some bios updates, 1,48v. The problem is that many options from a Z chipset motherboard, aren''t there in B660. The mobo is Asus TUF Gaming. So... Never buy B chipset paired with unlocked i7 and above
@@mattw.3724 Hey , I changed motherboard cos 1.385v was too much not needed voltage. I had to keep a certain bios to do that, latest bios had whatever I could possibly do , disabled. So I got stuck. I got my 13700k with 1.32v now , more stable (I can get down to 1.30 I think ). There is no way around a b660 or 760 mobo , it's Intel's way unfortunately
No joke I just helped someone on a discord I'm on because of this vid. They were having issues with a specific workload causing crashes and they didn't understand why, and turns out they just built new with a 14900K! Well default BIOS settings had everything cranked as you showed. Thank you, Jay. This was a huge help for them.
MSI Pro Z690-A DDR4 + 13600K , I originally only enabled XMP and assumed the CPU side would be fine with defaults but long story short MSI had the "CPU Lite Load" setting on mode 12 which is higher than Intel's voltage spec which lines up with Mode 9, I lowered it until the CPU became unstable which was mode 2 and raised it up to Mode 4 to be safe, my temps no longer hit 100C in Cinebench (around 86C) and gaming rarely sees above 72C (AK620 cooler). Also worth noting after a recent bios update MSI finally made Mode 9 the default as well.
Undervolting is one of the very FIRST things I do upon using new components, for both CPU and GPU. You can drop power draw, lower temps by 5-25%, all while losing maybe 1% of performance, if not _gaining_ performance if it’s thermal throttling and auto-underclocking. Even if you’re not having temp problems, it’s always good to set a minor undervolt because it can only do good at a stable setting.
I have a 7800X3D. It may score less than half of a 13900K in R23, but it will still shit all over it in gaming while using like 50W. So, I'm not too concerned. As long as it scores normal
Cinebench pretty much has no correlation to framerate in games unless the score is really low A 5700G scores higher than my 5800X3D in cinebench (15.7k vs 15.3k), yet in some games(notably Unity and UE4 games) the 5800X3D can almost triple the 5700G's framerate.
This videos was extremely informative, i always have struggled with BIOS and what to change or update. But after this I applied some your tips and my performance has increased and my PC is not running as hot. Thanks Jay!
Thanks for this Jay! I've been looking into getting either a 13600k or a 14600k (depending on sales), but the conversation around each chip overheating has made me wary. Happy to hear there's a way to bring temperatures down a fair bit (of course undervolting would help as well)
Im just building a desktop PC, i just had laptops in the last 15 years but now i need a powerful desktop. Im a bit confused, here i see he is limiting the wattage to the intel recommended 253W, on other vids i see people are tweaking just the volts and do some undervolting so which is better, limiting a 13900K to 253W in the bios or doing some undervolting? When needed i want the CPU to run at the highest frequency without throttling and losing performance.
@@Itsyesfahad because you dont need the Intel's processor hardware support for decoding/encoding, if the PC would be just for gaming i would have thought about AMD but i need to do some work too.
Thank you for your explanation jayz this is really help a lot !! It's crazy motherboard manufacture push CPU beyond the limits, i know is good for performance but for casual people like me who don't know much about this stuff it thought is just normal but is not, CPU brand should be need to cooperate with every manufacture motherboard brand to give the default settings for CPU itself so is not just came out from the box and shocked with ridiculous temps !! It would be nice if there is 2 option in BIOS which is option 1 is from CPU default settings and another option is from motherboard itself.
nice to see this is starting to get picked up by places. i went through a replacement asus z790 mobo, replaced a 13900k because i thought it was bad since new mobo didnt fix, experienced same bullshit with the 14900k that replaced it, all on "default settings" not realizing these assholes were just flat out unlocking wattage and amperage
I am a user of an i9-14900k and I decided to buy it because it was already starting to have a bottleneck with the grafic card, and after a few days of testing it went well but the temperatures were very high, like 80,90 and peaks of 100ºC. After seeing these results I didn't like them at all and I'm cooling the cpu with a Corsair iCUE H150i Elite LCD XT, and I thought what the hell is happening? And after doing a little research I found this video and I really thank you for this because I have gone from going to these temperatures to lowering them to 20ºC just by changing this simple option in the BIOS, and I really thank you very much for the video and for knowledge :) And to give an example in a game, rdr2 with everything in ultra does not exceed 60ºC :)
Z790 owner here & my 13900k just bit the dust, never overclocked & set asus optimized settings. Noctua cooler, & running heavy vms. Ran these settings immediately after this video on my new 13900ks & now asus released their much needed bios update to mirror this video with actual intel defaults. Honestly I am in no need to overclock an already heafty cpu & prefer stability & longevity of my pc. I hate taking my rig apart for some stupid settings to get a micro performance increase.
Honestly I am so glad you are talking about this. Going from an I5-10400 to an I9-14900k, I about spit my drink all over my monitor when I ran Cinebench on default settings and showed up to 5 different cores reaching 100C. After doing days of research on undervolting I found a subreddit where someone had mentioned the default bios settings on ASUS MB’s frying newer Intel CPU’s. So I followed his fix step by step and BAM! Instant 20 degree drop in temp without having to undervolt on cinebench. My score also went from 29000 to 37000 after multicore benchtesting.
Well it did! but I have no idea why it just randomly started like 3 months after using the ocmputer daily to have 0 stability like night and day. Anyone have any idea?
So I watched your video. Looked into the BIOS of my MSI Z390 - aaaand was absolutely clueless about what is what. Thanks for making me feel like a noob! I am gonna do a lot of reading now...
That's the worst explanation you've ever done. All over the place, completely unclear to anyone who hasn't been in and out of their bios and has extended their knowledge beyond that. Your speach is ranty, quite unstructured dialogue making things impossible for most people to follow, emotional and too fast. What is clear is that it's something you're clearly concerned about. Perhaps you could try that again but in a calmer more structured manner - that would be helpful to many of your viewers, and welcomed. Please have another go at this important subject, sit for half an hour before starting, and stay calm. Give more details for owners of the non Asus boards of what these options are called and where they hide. Thank you, it's important enough for a second video that's clear to everyone.
I have an ASUS board with an intel I7 and 360 radiator, I could not believe the heat it was pumping out, I took jays"s advise, dropped the voltage, I do not notice any performance loss, but the heat almost vanished, my fans now never really ramp up, and I also went back to windows 10 too, such a nice gaming experience again.
![You NEED to Change these 6 BIOS Settings now! 👉 [Lasts Longer, Faster, + Quieter]](http://i.ytimg.com/vi/1-7E3Je6qhc/mqdefault.jpg)
So the thing is intel programs a voltage frequency curve into the CPU. When the motherboard removes the power and current limits. The CPU will request insane voltages because it's not hitting the power or current limits. So technically the motherboard isn't feeding more voltage than the CPU is requesting. However if the power limits were being properly enforced the CPU wouldn't be able to request insane voltages.
So it's a safety for the ASUS/AMD issue we saw.
²À
This explains the entire video in one paragraph better than the video explains it in over fifteen minutes.
Does it means that undervolting is a necessary step to avoid overpower condition?
@@georgejones5019Yes and no. Intel is different in how it works because they have a set power draw curve. Specifically what ASUS is doing here is tricking the CPU into a false understanding of it's own state to get around Intel's hard coded limits. The CPU requests more voltage because it misunderstands what the MB is doing. In some respects it's worse than the AMD problem. AMD will mostly just let you send it whatever you want and it's up to the MB manufacturer to stay within the spec AMD gives them. With Intel they are purposely bypassing and taking advantage of a sanity check Intel put on the chip.
Yes, I agree that loading optimize defaults in the bios should be 100% safe and reliable. No overclock no auto adjusting.
but you're not describing "optimized" you're describing "fail-safe". 100% safe and reliable with no overlocking should be the default, but not the optimized default.
@@brenthauer8365Yet, there's a huge difference in between 'unstable' and 'destroys your hardware' which 'optimized' does not account for, neither in tradtition nor in meaning, because it has been there in the BIOS for decades and never killed your hardware before, and the word 'optimized' doesn't include 'destructive'. 'Optimized' never even came with overclocking in the traditional sense, it at best decreased RAM timings and set some little optons like AHCI instead of IDE and disabled energy saving features and whatever - but nothing even remotely endangering the integrity if your hardware.
And the most important thing about this is to acknowledge, that the vendors are doing this intentionally. They want their boards to be just the little bit faster than the others' boards in reviews, thus making unsafe settings the default. It would be very easy to simply add other settings like 'Dangerous' or 'Overclocked Defaults' to any modern UEFI setup, but they obviously just do not *WANT* to do that. They actually are misleading customers by lulling them into a false sense of security. Customers who pay many hundread or even a thousand bucks for a mainboard - for then being treated like idiots as a reward.
@@brenthauer8365 The problem is most, if not all, modern motherboards only give the option for "optimised" defaults. In this case being the only option they should be 100% safe and reliable without any auto overclocking regardless of whether they use the "optimised" nomenclature. But for argument's sake let's say a motherboard has both a "fail safe" and "optimised" option I would still argue that optimised should still be 100% safe and reliable with any automatic overclocking being kept well within the limits that both the CPU enforce and what the motherboard is capable of, it certainly should not be disabling any limits nor should it be trying to push those limits to the extreme.
@@brenthauer8365The point being made is that the optimized defaults IS the default setting out of the box, so these optimized settings are the ones you get when you buy your motherboard and first use it and also whenever you update your bios. The only way to not have these optimized settings is by physically changing bios settings. These should not be the default settings. That is the point of this video.
I had a 10th gen ROG board fresh out of the box on 10th gen brick it self during the initial setup because it was pushing “unstable” “”stock”” options.
Ironically I was building 2 systems and the other would crash **every time** installing windows.
Turned it off and fixed the issue.
Sent them both back. Eff that noise. They haven’t learned anything.
They want to sell you a new CPU not too long into the future. They have learned that a lot of people still use a decade old CPU and it Works! ...due to low silicon degradation, and that they couldn't improve sheer IPC that much. So the new business solution is to degrade (fry) the silicon so it only last trough warranty.
My toxic trait is having zero issues with my custom built PC and wanting to look for a problem to fix anyway
Mannn sameee
Ahh, a man of culture
@@death.r6 😆 🙌
Lucky, my computer has been completely imploding on itself
Same, it’s exhausting 😂
100%, this needs to change! I've been building for 20 years and I've never had so many out of the box issues with these BIOS settings. Please stop!
Asus is the worst for this... I had a sandybridge 2500k that got slowly died as the bios kept trying to overclock it... it's their buggy AI Tuner crap. Which is still at it on the 7950X with AM5, I've had to set an artifical voltage ceiling of 1.3 volts.
Hi! can you help me? I do that thing from the video but my is still the same settings... I need update the BIOS?
True 😎
To be honest, Asus is doing intel the favour here. U got more performance than what it should(intel power limit). Of course at a cost of power consumption. Which lead to other problem like fan noise, depending on ur cooling budget.
If all intel unlocked processors got benched by limit enforced.. its multi-core performance will get affected by quite a margin. And will look less desirable. Already at tough competition with amd..
I'd like to add this to the average viewer who wouldn't know this information as simple as I can put it, but Voltage (Volts or V) does NOT generate heat when applied to devices (CPU, GPU, RAM etc), under no load conditions 1.1v and 1.5v will give you the same temperature assuming no load as long as it's within specifications. (Including windows background processing)
However, current (Amps or A) are what causes heat generation when it flows and power is present (Watts or P = Volts x Amps) through the devices and potential damages can happen when current isn't limited or regulated using the voltage applied.
So what those motherboards are doing are basically removing the TWO PROTECTION limits of current and power, which is... Yea, very harmful for electronics...
To clarify, no motherboard should set default limits _higher_ than the CPU manufacturer recommends. If you have an A320 board and want to drop a 3800x in it, the board should absolutely enforce its own lower limits even if that means gimping the CPU. Better that than cooking itself.
mmm no
Tell that to ASUS and their 7950 X3D debacle.
@@grgspunkI remember dealing with that before upgrading to an MSI board last year, I stabilized the ASUS setup by giving a somewhat negative power curve to the 7950X3D.
@TheDivisionAddictMy AM5 B650 Elite AX at $200 is much more solid than my X570 Elite AC was at $200. Gen 5 on one port. I guess it depends on your buying segment. I wouldn't say that is the market as a whole.
I have an A320 board and been using A380 limits and have had no issues with it as of yet.
Oh my God... For the past 2 YEARS I have had issues with CPU temps under extremely minimal load. I have changed thermal paste, fans, AIOs and even bought a new CPU with no luck. After literally disabling one setting as suggested I am finally at 38c with 30% CPU usage. Thank you, thank you, THANK YOU
I'm going to assume its an asus board because its a 99.999999% chance that it is. Until they learn their lesson and it affects their bottom line they won't stop, it allows them to "win" benchmarks at the cost of your hardware. They know that the majority of "reputable" reviewers now only do default settings so they choose to do this on purpose.
Yeah a great JayZ video, have not seen this before on YT.
@@JohnSmith-ro8hk mine is a Asus! Never buy Asus again! They pretend to have good postsale support keep replying to my emails but with no real help!
You went about solving the issue ass backwards with only one datapoint
Noob
I just recently built a gaming rig/everyday family truckster with an i7 14700K and couldn't understand what the deal was, went from a Cooler Master 212 Halo to a Corsair H100X AIO to try and wrangle the outrageous CPU temps while playing Helldivers 2 and Arma III. So I wasn't losing my mind, it really is a voltage/heat issue stemming from the MOBO. Gonna dive into BIOS when I get home from work and investigate. Keep it up Jay!
Undervolting is the new overclocking. How far can you push it down?
The answer is 1/12th
about 350.
new? been doing it since 2007 :)
@@DC-te1gwgenuinely curious. Why?
@@SinisterSkyler Because you can get the same performance with less voltage, its like the reverse of overclocking, you just keep pushing it down until it crashes, then you bump it up a little and let it run at that voltage. This can keep your temps down by roughly 10-20 C just with undervolting alone.
Jay is an awesome technical content creator. With him, I improved my English listening skills, learned how to maintain my desktop computer, adjusted the tweeks of the components, and understood the synergy between them.
I heard that he is passing through a delicate medical situation. I wish him my best vibes to get over whatever condition he is in. Human being like him are what society needs to understand the meaning of "greater good" for being empathetic to others and helping others.
Thank you Mate!!
thanks2, glad to be part of your success, still its hard to keep positive mentality with this weirdness, fact is nobody give any $ which nobody cares, and if they still argue w classic 'afraid this that' then they also dont care abt me
@@iikatinggangsengii2471 tf are you yapping about?
🎯 Key Takeaways for quick navigation:
00:00 *😓 CPU and motherboard default settings can cause overheating and throttling issues.*
00:41 *🧭 Motherboard manufacturers apply optimized defaults that overclock CPUs and increase power/voltage limits beyond Intel's specifications.*
01:23 *📝 The video aims to explain why some users experience high CPU temperatures despite adequate cooling.*
02:19 *🔍 The issue is specific to Intel CPUs, as AMD CPUs are not pushed as aggressively by motherboard defaults.*
03:30 *⚠️ Enabling "Let BIOS Optimize" sets higher power limits, amperage, and voltages than Intel's specifications.*
04:42 *📢 Motherboard manufacturers should load Intel's default limits out of the box instead of aggressive overclocking settings.*
05:53 *🔋 Motherboards provide more voltage than the CPU requests, leading to higher temperatures and power consumption.*
07:45 *💻 High idle voltages are normal for CPU stability, but load voltages and temperatures are concerning.*
10:22 *🔄 Resetting to Intel's power limits results in lower temperatures and voltages while maintaining performance.*
14:07 *🚨 If experiencing high temps despite adequate cooling, check and reset motherboard settings to Intel's defaults.*
Made with HARPA AI
Until a company feels it in their wallet, they won't change it.
Totally agree it’s all about consumer power but we need to work together for it to work.
So general consensus is to deregulate, but the solution is to make them feel it in their wallet where they also get a bunch of money through B2B deals anyway…. Got it
This might work with game developers but it doesn’t really work in the main space where these companies have diverse portfolios.
Oh you’re not buying their motherboards? Just focus more on servers or something.
@@methane1027 20 years from now, it will be boutique in the west to buy a custom built salvage motherboard. The only thing stopping that will be anti-repair laws
Funny, that same perspective addresses 3rd wave Feminism too
@@poopingwhilestanding5801 possibly, but I beg you and everyone else to remain on Jay's topic of dumb mobo issues.
Just built a 14900k. Posted Aida64 stress tests to "it humor and memes" and had peeps concerned abour my 100c spikes. After your vid, went back to the BIOS and sure than shit had the defaults on. Enabled the Enforce limits and voila didn't go over 87 on a silent fan profile and 82 on max fan. Posted the results to Twitter and gave yiu a tag. Thanks a million for this vid.
what mobo? and did you change anything specific? also just built a 14900k first thing i did was cinebench and almost instantly go to 100c
@@zackregansounds ASUS Z790 wifi ii. Lian Li Trin Galah II 360. Contact frame.
I bought a Alienware r16 with the 4090 and 14900, it’s running hot too on demanding games it constantly is at 87-92. Smh I’m new to pc gaming. Scared to mess with settings like this
@@donjuan8124 87-92 on a PreBuilt I think is pretty good. Not sure what kind of overhead you have with the Dell Bios, but Prebuilts are know for spectacular air flow. I avg 87ish now OC'd to 6.2. Fan settings are on quite profile as well. So I'm happy with it.
@@jasonrichard5752 so I shouldn’t be worried about the temps like that? I called support and was told anything under 100 is good to go
This is amazing. Just built a new system with a 14900k and was baffled by the issue you described exactly. I'm not experienced with overclocking and voltages etc so I was a bit stumped as to why I was maxing out my temps then as soon as I took off the load dropped to like 30 degrees. Thanks so much for the vid, the timing was superb!
I don't leave comments to often, but WOW. I have been having this issue with my i9-14900k and had no idea how to fix it. Repasted and replaced my cooler multiple times. This simple setting on my ASUS MB solved it right away. Max temps now are around 85C during a Cinebench Multi Core test. !!!THANK YOU!!!
what scores are you getting? just out of curiosity
did u disable ASUS Multicore Enhancement and what else?
Some manufacturers are worse than others for high voltages, Asus may be guilty.
Did you change any other settings or only the settings in this video because i have the same specs as you and i followed the steps in this video but still getting 100c. Replaced the cooler and still same issue
@@fady12210 same as you. Any luck ?
THIS!!! This was me! I spent MONIES on making my first custom water loop, and I thought i had done it completely wrong, because every cinabench run i was getting thermal throttled... I could not figure out what i did wrong so tweak i did... and Boom its under control... pulling 350W at max settings on my 13900K at 96C 6GHZ all cores... so yeah my water loop works... it was the no power limit causing the issue.
Could not have fixed it without Jayz help.
Thank you Buddy!
I need your help. After disabling Asus MultiCore Enhancements, I only get like 5200 or 5300mhz on p cores when running cinebench. Also the current cpu core/cache current limit is 500 Ampere. According to Jason it should be 360 amps. What am I missing here?
You need a bigger or second radiator and/or a better water block. My AMD 3900x water-cooled stays in 50s to mid 60s C peak under full load. I have a custom loop with a good water block but an insane radiator - the Alphacool Nexxxos Monsta 560mm. My case actually supports two of them but I only use one for my CPU and 3080Ti. That radiator is rated for over 900 watts of cooling. I am currently planning to upgrade to AMDs Ryzen 9950X or 9950X3D when they come out and at least know that my water cooling loop which I designed and built with my 3900X upgrade will be sufficient for many more years to come. I've always believed in overbuilding on important aspects like cooling and power capacity. It makes for a much more stable experience.
I’m blown away that you would try to guild a custom loop without knowing the basics. I’m glad he was able to help you but this is not rocket science.
@@Fendera1you could start by asking the question in an intelligible manner. No one can see what CPU you have, what cooler, what motherboard, etc. no one will take you seriously if you can’t be bothered to provide that info.
@@KDarkmoon1 A 3900X is nowhere near as hot as a 13900K.
holy crap, this is literally a breakthrough video for me. I have been struggling with crazy temps for a while, and even went through all of the trouble to figure out how to UNDERvolt a i7-12700, but i didn't realize that hitting DEFAULT was actually over powering my processor.... I thought I was going to base settings by pressing defaults, but obviously not technical enough to read through the menus to figure this out. I have just found this video after changing my thermal paste, moving around my fans, considering replacing a 240mm Arctic Cooler Master II, this was the fix. Thank you!!
The only 2 things that will cause companies to stick to default limits are 1) Everyone refusing to buy boards that ignore default limits, or (more realistically) 2) Intel telling them to knock it off or they don't get to make LGA whatever boards anymore.
You should definitely show the AMD side. Although we're a small group of people, we would like to be careful too!
only thing that might be an issue with amd is the X3D chips
Oh we're definitely not small and we also had a fair share of issues like too much soc voltage frying our cpu.😂
Once you know how to undervolt it’s so easy.
@@mangatom192 That wasn't an AMD problem. MB problem (ASUS)
There is something called the amd bulge. I think it was on the asus mobos but not 100% sure.
Jokes on you. They intentionally did this to force you to replace your pc and make more sales. One way to boost sales in a recession 😂😂
jokes actually on them because now more people know about it and they can just switch their bios settings so it doesnt cook their cpu.
If possible, an equivalent video on the Ryzen 7 CPUs would be very interesting and helpful in making them run more efficiently and cooler.
And/or Ryzen 5
yes please
Look up ECO mode and how to set up PBO values for your CPU if you're using and X-version Ryzen. ECO mode will reduce your temps at the expense of speed, but if you're just gaming it's not noticeable. Personally, i'm okay with an avg of 1-2 fps loss to go down to 83 degrees under load vs 90 by default, but most of the time the fps loss is not even there either, so it's jst a win-win.
That one's not on the mobo manufacturers, but on AMD. If you want to be extra thorough, you can try undervolting along with it, but ECO mode itself will do just fine. If you're not on an X-ver Ryzen, you don't need to do anything, but you can always try undervolting.
All of my amd is running cool, that is why I would not own an intel junk
There's not such an issue on AMD CPUs (AM4 at least) you can play with the curve optimizer and PBO to get quite good gains in both performance and temperatures.
I wish this video had come like 2 days before. I literally spent a couple days off trying to figure the Asus settings on my wife’s 13700k.
I did the same thing this week.
Couple of days.😮 Of course all of us are nerds and like to make experiments too..
Thank you for this helpful video!!!
I just built a new PC last week with the Asus Rog Strix Z790-e Wifi II and used the Asus Load Optimized Defaults. When I ran Cinebench my CPU was running at 99C!!!
I then changed Asus MultiCore Enhancement = Disabled- Inforce All Limits and now the max CPU Temp is 72C.
Thank for keeping my new build from frying itself!!!
nice im about to do the same
Bro!! I made that one change and when gaming… went from 80.90C to 60sC. OMG!!! I was wondering if it was my water pump, etc. definitely saved my CPU’s life….
Man I think this trash mobo killed my CPU. These fucking things are getting too complex to build. I'm never buying ASUS again.
After four months of trouble free usage, the system started to lock up after a minute. I definitely used default settings...
@@jeezusjr same thing happened to me. Bought a new Mobo, CPU, psu, cooler, and ddr5 ram to upgrade an old system. Hopefully get some of the $ back from returns Lol
I also find it insane that because of these settings I have had multiple customers come in with Asus motherboards and a 12900K or 14900K and the Cpu was unstable and they couldn't figure out why. It would crash with strange errors, one of them even claiming it was a memory error, when the only issue was that the CPU couldn't handle the power that was being forced into it and so it was crashing.
i went out to shop for 14900K with z790 last week, and i met another customer who made the same build 4 days ago. he came back complaining his i9 going 99-100 deg in 5 seconds, and wont come down even at low loads. according to him, his super expensive AIO was at fault, so he just asked the store manage to replace his cooler, lol. I told him, its not the case, just undervolt the cpu and adjust till you achieve max freq and low temps.
conclusion: he still went home with a different cooler 😅🤣
@@deepak_nigwal same thing happened to me i got a z790v board with a 12900k i got a water aio cooler. Took it to microcenter and they installed my cooler wrong they.i was pissed but they fixed it and it worked but i be download msi afterburner everything fine but my temp says 120° cpu idol .i turned off my pc checked in BIOS and my temp was a cool 50° . But on msi days it was cooking steaks but my boss said my cpu is fine ????
@@deepak_nigwalthe mistake they did was they didn't adjust the aio coolee bracket so I'm thinking that was the problem bcuz I'll play cyberpunk and my pc runs fine for hours and after i check my bios ... Under load was 60° ....
what did u change then?
In some cases I just had to turn off MultiCore Enhancement. In others I actually had to underclock the CPU by a 200 to 300 MHz and then it became completely stable. I had a computer in the shop last week that was throwing memory errors when launching games and there was no issue with the memory at all. The only issue is that the CPU couldn't run stable at the speeds it was binned for in certain circumstances. So dropping the core by 300MHz, basically running it as a non-K CPU was all that was needed to make it 100% stable. Before it was around 80% stable. @@byFraze
Thank you for this, I just done this with mine and I see a big difference in temps.
Thanks for making this video Jay. My 14900k was hitting 100c on multiple cores whilst gaming and benchmarking on "default" settings. Even though having a high-end 360MM AIO.. Once I disabled that setting I'm getting a stable 81-85c and the performance loss is only 2.1%.. To think for all these months i've been torturing that CPU for an extra 2% performance...
which setting did u diasble? i watched this video 3 times and cant find the part where he tells what to turn off..
Same problem here, I can't see where in the video he makes the correction
So glad i have the 7800X3D... Just set the curve on -20 and works perfect.
Ya man. I went in to Ryzen Master and just clicked on auto overlocking and just let it do its thing for my 7600. Came back with a -30 for me.
What does -20 mean? I run a stock 7800x3d
I clicked on the video thinking AMD board partners had gone back on the fixes from the first batch of melting am5 socket cpus
@@Grillhandle they're talking about the PBO curve
i set it to -30 on my 5800x3d, insane how much the temps dropped after that. still need to wait and see if it's stable but so far so good. it also performs even better now.
This happened with my 10700 on an MSI board. For the longest time I thought I had something thermally wrong. Chased it for a while with no results, then one day read about how motherboard defaults are ridiculous and, just like in this video, the motherboard defaults were nowhere NEAR the normal Intel limits for the CPU. Switched to Intel's base recommended defaults for the CPU and it instantly solved the thermal issues.
Even aside from thermals, not only does it confuse consumers and waste time/money chasing problems not caused by the consumer- it wastes a ton of extra energy too - so thermals and energy skyrocket with extremely diminishing returns. It's so frustrating that this has become standard practice.
This but 10700k and two ASUS/ROG boards.
One crashed during UEFI update time because unsafe thermal/power limits. The other crashed every time installing windows.
The CPUs overclocked just fine tho.
Returned both boards and went with a different brand.
Hard crash during UEFI update = dead board.
BIOS Flashback thingy didn’t do jack.
Hi @neonvoid666 , actually just got through building a new pc with an MSI MPG Z790 Edge Wifi. So how would I go about setting the mother board to these settings you did. 13700k, 4070ti, LS720 SE Digital aio cooler. I would really like to do the Intel defaults. Cheers.
Can you share what i need to chamge i have a i9 13900kf? But i dont know what to change
What settings did you change?
This is excellent information! When the max power limit was set to 4096 Watts (by MSI) instead of 154 Watts for my 13th gen CPU, gaming was extremely difficult with stutters and input delays of atleast 500ms, even walking/running in game felt difficult, counter-strafes never registered. After changing it to Intel's default 154 Watts, gaming is not painful anymore. Please make more videos on BIOS settings - Disabling C-States, Intel Speed Shift, Speed Step, Secure Boot.
what exactly did you change? i got the same problem with inputlag.. maybe this is the final thing to test if it fixes it
Just built a new i9 14900kf system, with an Asus motherboard, 360 AIO, was having terrible temps, changing this setting really help. Really appreciate your videos.
If you're using a MSI motherboard, the power limits are auto set depending on what cooler you selected in the bios. If you select water cooled, it sets the max watt to 4095 like shown in the video. Change it to box fan even if you use an aio if you're having temp issues.
TRUE
What about intel turbo boost? Disabling does help or not?
And what options i should disable @ msi bios, thank you
@@ugurbaytar Turbo boost kapatırsan performans düşer.
@@HuCuRuS hocam kapatmadan da boşta bile 100 derece zaten öyle olunca da performans düşüyor artık sıvı soğutmada var bir problem diye düşünmeye başladım
Great vids, always a joy to watch. I'm sure plenty benefited fully from it I also did but to an extent because I own an MSI laptop and their Bios is not exactly a walk in the park it's so so confusing, I really wish you guys will get the chance to expand it into the laptop segment too in terms of technical stuff like this vid where someone like me and I'm sure they're plenty because I looked and couldn't find anything in that regard when it comes to tuning an MSI laptop with an unlocked CPU variant through the Bios.
Let's talk about it at 0:38
Jay, please let it be a 45 minute video. I could listen to you vent all day long. Thank you for all you do!
You are a life saver, I used ThrottleStop to address the motherboard settings issue, it was not enough, I still had power and temp fluctuations, even with base power PROCHOT lit up, entered the FiVR window, and set everywhere I saw 511.75 A, 307 A and the fluctuations disappeared. However, if I want to change the power limit from 125W to 253W, it doesn't apply it in ThrottleStop, Prime95 remains with 125W power, I entered in XTU and asked to reboot, before I did not need to reboot, I have to learn more about ThrottleStop looks like. I got a B760 mono because the Z790 one was 70% more expensive in Romania.
Thank you for releasing this video. I bought an R9 7950x like 8 months ago with a 360 aio... and it idles at like 60c... and doing anything that takes cpu skyrockets to 100c. Tech yes city mentioned undervolting because of how insane they are cranking the settings to compete. I dropped my voltage from like 1.2-1.4 (don't remember stock voltage) to 1.000 and it dropped to 35ish c idle the second I click the apply button. I have an msi motherboard. This just confirms I need to adjust my bios. I did absolutely nothing in my bios except maybe cranking my ram up, but I don't remember. Time to learn bios and fix it....
just set your cpu to 105w eco mode.
Pretty spot on and important. Went on a MSI Z690 from a non K 12600 on a decent aircooler to a 13600k and loaded "defaults" and at a sudden temps skyrocketed under stress. Couldn't for my bare @ss figure out what was going on but knew the new 13600k had some higher TDP so went and installed a decent 240 AIO instead. Well, the temps was still very high but not throttling but the TDP still "out of league". Then I started catch the info about "undervoltage your rig" etc and one of the most important settings on MSI was the "Lite load" settings that currently were at "12" I finely read the voltages and saw they were way too high! Played with the "Lite Load" and ended up with set it to "4" istead. WOW what a difference! Now I finally have the normal TDP 0f 180'ish watts and normal temps and still doing speeds as on the box and still rock stable! Shame on yours motherboards manufacturers litterally cooking the CPUs for marginal gains!
The latest MSI bios is a lot better about this, the BIOS my Z790 came with was pulling 220w on Cinebench23 with a 13600K but with the latest it's around 180w. I used the CPU lite load setting to lower the voltage a bit and now it's 150w and all cores can run in full turbo full time. I can get it down to 120w but I bumped the CPU lite load setting it back up by two increments to be sure of stability.
This video saved my life !!!. I just got 13700F and when i play BFV my cpu was hiting 100 degrees. I even decided to buy some AIO and other stuffs but when i disabled the ASUS Performence Enhacment 3.0 my max temp is lowered to 70 degrees. Thank you so much!
literally just went through this. Ran Cinebench with my 13600k w Artic Freezer II 240mm, was hitting 100*C almost instantly. I bought a contact frame, installed it and had the same result. Did some googlin, and found that my MSI mobo has a setting called CPU lite load, which was set to auto by default. Changed it to setting 10 (from 12) and am now maxing out at 90*C in Cinebench with more consistent clocks, and score within margin of error of previous. I could probably tweak some more stuff to improve it even more but it doesn't really matter for the day to day stuff, and it does idle/run games cooler now with the contact frame.
U can easily go to 5 , and still have same perf and about 10c less .
JAY YOU ABSOLUTE GUN!
I built a PC about 4 years ago and used auto for most settings, I would often hit 100C and actually overheat even though I used a liquid cooling AIO.
I changed some settings in line with this video, and I'm now no longer going above 55C or experiencing fans maxing out RPM for load
Great stuff!
Same as you, my 13600K going from 100°C to 55°C in gaming...
Which settings boys?
Any performance decrease doing this?
Bro wich settings? Im currently experiencing this myself, cpu just goes to 100 celsius instantly while being IDLE, I have a watercooler aswell and I have this pc 4/5 years now and never any heating problems, what should I do?
@@CrisperFN if it's instant you need a good clean out and repaste. That's too fast to reach 100c
I needed this as I've just built my PC after 8 years of having an OG beast. I appreciate this channel a ton.
This video just help me out so much. For almost a year I've been fighting the temps on my 15 13600k and I just thought it was cooler/fan. After doing some of the suggested changes mention in this video my temps are FINALLY normal. No more 99c cpu @ 20% usage. Thank you and Keep up the good work guys!
The "extreme tweaker" mode is actually meant to liquify all of the copper in the motherboard so you can collect it and go sell it for some more meth
I heard that wasn’t true by a lot of people
Its almost like everyone needs to take a hardware generation off and just work on efficiency and having things work properly. Imagine if today, we had the performance we have but our CPU's still never really went over 60c?
Over 4,000 watts?!?!?! That motherboard may as well just come with a fire extinguisher and a map to your circuit breaker panel
The power supply would probably blow up before your motherboard
There will be no over nine thousand power level and no 1.21 jiggawhats. It just appears to be a 12-bit value (not uncommon in ADCs and PWMs) which takes values from 0 to 4095, the latter value effectively meaning that the motherboard shouldn't be bothered by what your cooler is capable of. The same goes for the current limit of 511A. It's 2^9-1. These capablities may be used by very special, very small systems with small VRMs and heatsinks (think embedded, industrial, battery powered, aerospace, military), but are of no concern on regular mobos with 8 or more phases of voltage converter which is actually obscenely huge. I run my system with both settings maxed out for more than two years and it still didn't blow out the magic smoke.
We need to make AIOs with one of those ceiling sprinklers build in.
A CPU has impedance, it's impossible to reach those numbers.
It doesn't use 4000w. That just means unlimited. For 13900s it's about 300-350, for 14900 pushing 400 depending on how fucked the defaults mobo is.
Excellent video. I think one of the reasons this is still happening in 2024 is because it is simply not being talked about enough. So many thanks to Jay for doing so. Why having your CPU constantly above limits and thermal throttling to protect itself as a standard default setting is beyond me. The point being made is that it is fine to have this setting available should you wish to use it, but it should never be the bios default setting. For those concerned with performance (gaming), I would encourage you to actually test this. You will probably be surprised by the results.
Hey thanks a lot for the information and honestly I did noticed a huge improvement with just disabling that feature 👍🏼
Years ago I had a Intel I7 7700k and a MSI Mpower platinum z270 motherboard. Running the bios factory settings gave me a idle temp of 80c. Found out that the core voltage was set to 1.4v out of the box. I dropped it down to like 1.2 and idle temps went to like 30c. No real performance changes as i dont over clock or play around to much with it, MSI released a Bios update that changed the default to 1.3 as clearly they identifed this problem. I still run it as 1.2 and have no issues. This has been issues for years.
Thank u for sharing this I must check my own now 😅
I did the same. I have the i7 7700k and a gigabyte Motherboard. My idle temperature was about 80. I decreased my vcore voltage to 1.1 and solved the problem.
Thank you! This is such an important thing to get out there. I feel so bad for people building their dream rig and then feeling like THEY did something wrong because of high temps due to motherboard vendor default OC profiles.
High temps aren't a problem on current gen. CPUs are now designed to turbo until they have to reduce clock in order to stay under TJmax.
This is a good thing, you get more out of your CPU doing so, and your cooler is more effective when your CPU is hot (deltaT and all that).
The old school users who are used to trying to manually keep temps under a certain level are having a hard time adapting to this idea.
@@jttech44 overvolting makes the cpu heat up faster and throttles it more.
I'm using an MSI z790i but their latest beta bios states "that it has replaced their own system power settings and replaced them with Intel defaults but users can still optimize system performance with alternatives from MSI." At least MSI has taken a step in the right direction, I typically shy away from beta versions unless it has a feature that I need.
Have they released a stable version that implements intel defaults at stock for your MSI z790i or is it still in beta?
Because proprietary stuff. Why share money or care about the consumer as long as you get yours? Greed and money ruin everything.
Thats the whole world today unfortunately
if money ruins everything you can give all of yours away.
Except I don't have any.
@@soccerguy2433I stand with blockbuster
@@Dave7heRaveIt’s that way now because are taught only to care about materialism
I literally finished building my pc 2days ago, this makes SO MUCH SENSE!!! Honestly thank you!!
Gosh! I spent so much time back in 2016 trying to figure out what was wrong with my i7-7700 (65W CPU) going way above 90°C all the time... I even bought a new cooler thinking that my Noctua cooler wasn't doing its job. I damned Asus when I found out months later that it was MultiCore Enhancement's fault... I didn't know they were still doing that after all these years. They must stop. This should be optional, not default. That video was necessary.
If you have an MSI motherboard change Lite Load from 12 (default) to 9, which is the Intel recommended setting. It makes a big difference in temps and scores.
Mine was defaulted to 9 already
Yep, ended up with could set it as low as 4 w/o issues. WAY better temps!
Mine is also 9 but with 12700k, in previous bios I should set it to Mode 1 to be like I was when I first bought the motherboard... Now my normal is 7. You should just take a fast CPU bench with cpuz and check your score. Lower mode after till you see less score and then up it by 1 (or maybe 2 to be sure that no speed penalty exist)
@@haies09 That’s interesting. Mine was 12 auto. I just did a search and found reports of some people saying their auto is 12 and some where they say their auto is 9. I guess just always check to make sure.
@@CindersTV being I don’t hit my base clock speeds and ain’t thermal throttling, maybe I can increase it to 12?
To further this, go to Global Core SVID Voltage. Offset Mode Sign to Negative , Additional Turbo Mode to Auto and Offset Voltage to 0.10000. It will help to properly under volt your processor without hitting performance.
Video title should be "Motherboard Default Settings ARE actually COOKING your CPU".
Jayz too humble on the fax
Congratulations on reaching 4M
But if they don’t artificially inflate their “defaults” as faster than the competition, how can they claim their products are “better”? And if they don’t reduce the lifespan of their products, how would they make more sells if a system takes 3-5 times longer to fail?
@JayzTwoCents thank you so much for posting this video. I've been having horrible crashes, including some blue screens, with my newly built homebrew system. It's been having random crashes when under heavy load (usually cyberpunk or bg3), and hardware diagnostics couldn't find anything wrong. I disabled this setting, and I now haven't had a crash in two days. You are a lifesaver!
years ago and i am talking 25+ years ago they used to have load default settings and load optimized default settings. and the load default settings was the factory settings for that cpu. and the load optimized default settings was over clocked settings for that cpu based on motherboard manufacture settings for that cpu. they should be still doing this and i do not know why they stopped doing load default settings
There's no default settings based on the CPU now? Then what the hell are they based on? This is scary.
Exactly my problem ..... I got Asus maximus hero x MB with delided 8700k on custom water loop and its hitting 100c and I was thinking i got bad cpu ..... started to look for replacement cpu and stumbled today on this video. Soon as I get home from work I will test the setting.
Please share your findings!
Yes, please update. I also have this same board !
The crazy power limit is something I noticed immediately and quickly set the watt limit to 253. My 13900k stays at 67°C on cinebench r23. I also found gaming was noticably better with the 253 watt limit(as in movement/feel). Running on a MSI Tomahawk z790.
How do you do the 253 watt limit?
I loaded my BIOS back to default and did not fix the issue, but I was able to find a fix within Windows in power management and it resolved the issue with the CPU.
What setting was that?
@@Chris256 You will need to go into control panel, then select change plan options. There's an option for change advance power setting, from there you will select (Processor power Management) If Maximum processor state is at 100 switch to 99. Issue will be resolved.
@@Chris256For some 13 & 14 gen you may need to change it to 50%.
@@pjm7482 It will turn off the turbo boost doing that. Its a shame that we even have to do it bc some motherboards don't even let u change anything. And we missing out on performance bc we can't have turbo boost on without cooking up or cpu. I guess u could make your own powerplan to somewhat get out a little more performance without it hitting over 1.4 volt .
i don't understand, are the showed tips, e.g. "Gigabyte: Tweaker--> cpi upgrade: Default" your recommended settings or the settings that cook our parts? if they are the bad ones, which one should we chose?
I don't know how to pin a comment, but this one needs to be seen by Jay. speculate the video poster is the one who has that control. Confusing as he goes so fast.
Update: i downloaded cinebench and checked the temperatures in HWMonitor. The settings in Jays Video (e.g. for Gigabyte: Default & Auto/ enabled) instantely brought up the temperature to 100° C when i staeted cinebench. It didn't even take 10 seconds. So i guess the settings shown in the video are the bad ones, but i have no idea what to change, since in the gigabyte bios, the only other option apart from Cpu Upgrade: Default is "gamer mode" or "turbo" - which i honestly don't want to try out, since my flat is not flame proof.
@@zsedz 🤣
pretty sure you need to set turbo power limits to Intel POR or else Enabled + your own limits
The gigabyte manual says Auto is for intel defaults, but i think its incorrect and is actually gigabyte defaults and they edited the video based on that manual
@@ray_bromanook i will try this as soon as i get home, thank you very much for your help
OMG You saved my PC. I could cry, I have had this issue for years and even gave up with my pc for 3 years because of this issue. You found the fix, thank you so much!
So I went and checked and sure enough this is how my 13700k on a Asus board was set. I just assumed the default wasn't a oc. Dropped my temps in a cine run almost 10c lol
Undervolt by .4 and it will drop even further and run faster as it won’t thermal throttle
@@timothygibney159It didnt drop the clocks at all it sat 5.3 the whole run at 87c. Im fine with it how it is but i always wondered why it ran as hot as it did and this seems to have fixed it. Before it would shoot right up to around 95 96 and sit there i dont remember the scores i got last time it was awhile ago when i ran one but the test run i did after the change was still a little over 29k.
@@Premier024 mine I have to to prevent thermal throttle. It only has 95sp rating. I can hit 38k on cinebench now after the undervolt since it would down clock previously
@@timothygibney159 That's crazy 38k on the 13700k. Tbh I'm perfectly happy with it just getting what's it's supposed to with better temps I'm GPU limited almost all of the time at 4k as it is but the less the fans have to ramp the better.
@@Premier024 oh it’s a 13900k 😂. Close to the 13700k but shoot most other 13900k even with an overvolt can do 42k and 6.1 p core on at least 2 cores. I have 57 on 2 and that is it . I assume a good 13700k could hit 35k on cinebench easily if you adjust the volts and get at least 1 core at 6 gjz or 2 pcores at 5.9. I only light game old mmos and use this for virtual machines as a more cores are better for my use
My spouse is using my previous rig, with an Asus ROG Maximus VII Gene, a 4790k and a 980 Ti (had a 1080 but it ded). The Asus BIOS wants to run its adaptive boost, but honestly it cooks the CPU. Yes, it has been re-seated and re-pasted. It's gotten so bad it will no longer post with XMP enabled, when it used to run fine. Turning adaptive boost and multicore enhancement off is the only way that rig is still safe for gaming. I know that's older than the start time you mentioned, but it feels so related.
by re-pasted, do you mean de-lidded and the die re-pasted? don't forget the 4790k has paste as the TIM, not solder, so by now it probably needs replacing with something extremely viscous like thermalright TFX (the paste that's so viscous you have to warm it up so it even spreads. make sure to spread and have the sides covered too), or a honeywell PTM7950 pad if it's thick enough (the phase-changing cooling "pad" that GPUs are using on-die now instead of paste. it's only beaten by liquid metal but is much safer, but still pretty expensive. usually 0.3mm thick)
edit: misspelled viscous
@@glebglubNo, just reapplying TIM. I've never delidded a CPU (my first computer had a lidless AMD CPU but that doesn't count).
@@Efreeti might be worth looking into then, just like re-pasting an old GPU that still has life in it. there's minimal risk with a de-lidder tool (as opposed to the vice or razor blade method which killed mine from digging into the substrate by accident), but whether it's worth it to you both money and time-wise, that's not my call to make
A friend of mine had stability issue and bsod with the 14900k. He solved it by putting the cpu tdp and core amps at the value reccomended on the intel's spreadsheet instead of the automatic mobo value.
Hey Jay, please consider making a video where you test multiple different thermal pastes! Noctua NT-H2, Thermal Grizzly Kryonaut and Kryonaut Extreme, Arctic MX-6 etc. Especially because new pastes come out every couple months, it would be interesting to see which ones are the best!
There are videos like that on RUclips and last time I checked it didn't matter realistically as long as you use one of the top brands.
It doesn’t matter enough to worry, just pick one
Already been done. Thermal paste doesn’t matter
You dont even need thermal paste. ;)
There is liquid metal and there is any decent brand of everything else within margin of error.
Thanks Jay, my pc randomly shuts off while gaming, and it's brand new, so im going to try this seeting you just showed, thanks again
Sounds like anti-overheat tripped.
Always nice to see that my i7-8700k is still holding up well after 7 years because I adjusted these settings and delidded it as soon as I got it. I only had to reduce my 5ghz overclock to 4.8 recently to maintain stability.
Make sure to double check your ASUS MOBO settings. Because even though I set limits it was still running at 511, 4095 and 4095.
Correct. I set "enforce limits" and even so, I needed to set express limits on the voltage. I have an Asus Tuf z790 - Wifi Plus.
I have this exact problem with a B660 and 13700K. I had to workaround to get 1.38v, whereas itself gave 1.61 (almost melting socket voltages) and after some bios updates, 1,48v.
The problem is that many options from a Z chipset motherboard, aren''t there in B660. The mobo is Asus TUF Gaming.
So... Never buy B chipset paired with unlocked i7 and above
same here, Asus Prime B760-PLUS with a 13700K and its been a nightmare. love the vid, but I cant change those settings.
@@mattw.3724 Hey , I changed motherboard cos 1.385v was too much not needed voltage. I had to keep a certain bios to do that, latest bios had whatever I could possibly do , disabled. So I got stuck. I got my 13700k with 1.32v now , more stable (I can get down to 1.30 I think ). There is no way around a b660 or 760 mobo , it's Intel's way unfortunately
No joke I just helped someone on a discord I'm on because of this vid. They were having issues with a specific workload causing crashes and they didn't understand why, and turns out they just built new with a 14900K! Well default BIOS settings had everything cranked as you showed. Thank you, Jay. This was a huge help for them.
I have a i9-14900k, I applied what you did this video, I went from 100C to 65C max load and 29-32C idle
Thanks Jay!
Yea….i have to do this myself.
Make sure to undervolt it by 0.05
It will lower Temps and increase performance
What cooler do you have? I got very little change on temps.
@@dlmac I have a Lian Li Galahed II
MSI Pro Z690-A DDR4 + 13600K , I originally only enabled XMP and assumed the CPU side would be fine with defaults but long story short MSI had the "CPU Lite Load" setting on mode 12 which is higher than Intel's voltage spec which lines up with Mode 9, I lowered it until the CPU became unstable which was mode 2 and raised it up to Mode 4 to be safe, my temps no longer hit 100C in Cinebench (around 86C) and gaming rarely sees above 72C (AK620 cooler). Also worth noting after a recent bios update MSI finally made Mode 9 the default as well.
Please where specific did you find this setting?
I found it. Is this lite load that thing what was Jay talking about?
you can do that too but jay talking about asus multicore, it’s called enchanced turbo in msi
Undervolting is one of the very FIRST things I do upon using new components, for both CPU and GPU. You can drop power draw, lower temps by 5-25%, all while losing maybe 1% of performance, if not _gaining_ performance if it’s thermal throttling and auto-underclocking. Even if you’re not having temp problems, it’s always good to set a minor undervolt because it can only do good at a stable setting.
Me with a 5600X getting around 11.5k not overclocked lol. Hearing these stupid high numbers feels insane.
I have a 7800X3D. It may score less than half of a 13900K in R23, but it will still shit all over it in gaming while using like 50W.
So, I'm not too concerned. As long as it scores normal
Cinebench pretty much has no correlation to framerate in games unless the score is really low
A 5700G scores higher than my 5800X3D in cinebench (15.7k vs 15.3k), yet in some games(notably Unity and UE4 games) the 5800X3D can almost triple the 5700G's framerate.
@@HMSNeptun thats the 3d cache talkin.
@@rustler08 Wanna try that in a Simulation game like Minecraft? Cause your processor will absolutely get railed by a 13900k
@@HyperionZero Sodium and Lithium: who the f*ck need a heater to get a nice performance?
What should I change on a MSI motherboard?
This videos was extremely informative, i always have struggled with BIOS and what to change or update. But after this I applied some your tips and my performance has increased and my PC is not running as hot. Thanks Jay!
Thanks for this Jay! I've been looking into getting either a 13600k or a 14600k (depending on sales), but the conversation around each chip overheating has made me wary. Happy to hear there's a way to bring temperatures down a fair bit (of course undervolting would help as well)
That's why I feel bad for the casual users.
Im just building a desktop PC, i just had laptops in the last 15 years but now i need a powerful desktop. Im a bit confused, here i see he is limiting the wattage to the intel recommended 253W, on other vids i see people are tweaking just the volts and do some undervolting so which is better, limiting a 13900K to 253W in the bios or doing some undervolting? When needed i want the CPU to run at the highest frequency without throttling and losing performance.
@@AexoeroV Just buy an AMD platform and call it a day.
@@Itsyesfahad i need intel quick sink for some editing and i can get the 13900k at a very good price
@AexoeroV It's up to you man for me, I would get 5950X or 7950X instead.
@@Itsyesfahad because you dont need the Intel's processor hardware support for decoding/encoding, if the PC would be just for gaming i would have thought about AMD but i need to do some work too.
Thank you for your explanation jayz this is really help a lot !! It's crazy motherboard manufacture push CPU beyond the limits, i know is good for performance but for casual people like me who don't know much about this stuff it thought is just normal but is not, CPU brand should be need to cooperate with every manufacture motherboard brand to give the default settings for CPU itself so is not just came out from the box and shocked with ridiculous temps !! It would be nice if there is 2 option in BIOS which is option 1 is from CPU default settings and another option is from motherboard itself.
This is my exact same CPU and Mobo brand and this video has been super helpful. Thanks Jay!
nice to see this is starting to get picked up by places. i went through a replacement asus z790 mobo, replaced a 13900k because i thought it was bad since new mobo didnt fix, experienced same bullshit with the 14900k that replaced it, all on "default settings" not realizing these assholes were just flat out unlocking wattage and amperage
I am a user of an i9-14900k and I decided to buy it because it was already starting to have a bottleneck with the grafic card, and after a few days of testing it went well but the temperatures were very high, like 80,90 and peaks of 100ºC. After seeing these results I didn't like them at all and I'm cooling the cpu with a Corsair iCUE H150i Elite LCD XT, and I thought what the hell is happening?
And after doing a little research I found this video and I really thank you for this because I have gone from going to these temperatures to lowering them to 20ºC just by changing this simple option in the BIOS, and I really thank you very much for the video and for knowledge :)
And to give an example in a game, rdr2 with everything in ultra does not exceed 60ºC :)
Thanks for this, just dropped my temps 35.7c by changing that, now the wife is telling me it’s to cold in my room and to turn off my portable a/c. 😂
Thanks capt two cents! Exactly my issues with temps and whatever else these default settings were doing are getting sorted now to how they should be
Z790 owner here & my 13900k just bit the dust, never overclocked & set asus optimized settings. Noctua cooler, & running heavy vms. Ran these settings immediately after this video on my new 13900ks & now asus released their much needed bios update to mirror this video with actual intel defaults. Honestly I am in no need to overclock an already heafty cpu & prefer stability & longevity of my pc. I hate taking my rig apart for some stupid settings to get a micro performance increase.
In Gigabyte bios i need to set Turbo power limits to Enable or Auto?
Maybe Auto
This has aged well.
Honestly I am so glad you are talking about this. Going from an I5-10400 to an I9-14900k, I about spit my drink all over my monitor when I ran Cinebench on default settings and showed up to 5 different cores reaching 100C.
After doing days of research on undervolting I found a subreddit where someone had mentioned the default bios settings on ASUS MB’s frying newer Intel CPU’s. So I followed his fix step by step and BAM! Instant 20 degree drop in temp without having to undervolt on cinebench.
My score also went from 29000 to 37000 after multicore benchtesting.
Whats the Gigabyte Aorus equivilent of this for the BIOS setting??
would love to know this as well
Getting mad BSODs recently. Cpu benchmarks instantly cause BSODs too. Could this fix it?
Well it did! but I have no idea why it just randomly started like 3 months after using the ocmputer daily to have 0 stability like night and day. Anyone have any idea?
Jk my cpu fried from asus default settings. Just got my replacement today and IMMEDIATELY set those settings!
So I watched your video. Looked into the BIOS of my MSI Z390 - aaaand was absolutely clueless about what is what. Thanks for making me feel like a noob!
I am gonna do a lot of reading now...
That's the worst explanation you've ever done. All over the place, completely unclear to anyone who hasn't been in and out of their bios and has extended their knowledge beyond that. Your speach is ranty, quite unstructured dialogue making things impossible for most people to follow, emotional and too fast.
What is clear is that it's something you're clearly concerned about.
Perhaps you could try that again but in a calmer more structured manner - that would be helpful to many of your viewers, and welcomed.
Please have another go at this important subject, sit for half an hour before starting, and stay calm. Give more details for owners of the non Asus boards of what these options are called and where they hide.
Thank you, it's important enough for a second video that's clear to everyone.
I’m still confused on what I’m supposed to disable for my gigabyte motherboard..
@@ShottaNIflheim same
@@ShottaNIflheim same...
I have an ASUS board with an intel I7 and 360 radiator, I could not believe the heat it was pumping out, I took jays"s advise, dropped the voltage, I do not notice any performance loss, but the heat almost vanished, my fans now never really ramp up, and I also went back to windows 10 too, such a nice gaming experience again.
Thank you for calling them out!
I want peace of mind when doing work on my PC that it's not going Blue Screen in the middle of my day.
Oh man I wanted this Uline packing table and monitor arm set-up for so long I love it.