I'm just happy to see these finally coming out. On a side note AI summary is getting wild. "Intel released a microcode update (0x1 129) for the 14900K CPU to address voltage stability issues. The update resulted in a slight performance drop in some tests, such as Time Spy Extreme (700 points), but no significant changes in most other benchmarks. The update also reduced voltage spikes during single-core workloads, potentially improving stability. However, the author notes that it's too early to say if the update fully resolves the voltage issues and recommends waiting for further testing."
@@Jayztwocentsit's the " Ask" button that shows up under the video in the default Android UA-cam app on my pixel 8 pro. Basically auto summarizes most videos. It doesn't show up on all videos, however I notice it now far more than I used to. Nearly all at this point.
They should also release a testing executable to perform stability tests. By the amount of data they have at this point, they should know what to look for.
Intel PDT (Processor Diagnostic Tool) already exists, they'd just have to update it with a test that explicitly produces errors if your CPU is affected.
@@styleisaweapon Lol i was thinking the same thing, they tried to reject RMA as much as they could, so why would they release a tool to identify stability problem ? That would only increase the amount of people asking for RMA 😂
>It's time to RMA that, however painful that process is gonna be I'm in the middle of a RMA for my 14900K and Intel loves nothing more than employing dilatory tactics. It’s been weeks of drawn out emails, taking photos of the cpu, sending invoices and receipts, and they’ll do anything other than help me. I would have gone AM5 had I known how badly this was going to turn out.
this is why i dont buy anything for the warranty. every company does all they can to not pay out so why bother with any amount not worth getting a lawyer involved on.
I was one of the early adopters to 13th gen intel. I made the mistake of going cutting edge while Covid was keeping everyone home, so I blew over 4k on a new 13900k and 4090 rig I built myself. I ran so many tests and diagnostics, tested thermals, edited windows configs, REINSTALLED windows like 4 times, and reflashed bios. I reinstalled drivers and even replaced my RAM and motherboard entirely. But no matter what I did, I kept hitting bluescreens and MANY game crashes that I just chalked up to be faulty 4090 graphics drivers that have yet to be patched into a playable state. Only just now, MANY months of pain and anger, am I finally finding out what happened to my build, and why everything went sour. Intel's quality control, and now PR trying to cover up the issues for the last ~5 months, has indefinitely lost them a customer from me. I will never be supporting Intel in their shady ass business practices again. Now that its been over a year of having this PC and I've ran into too many headaches to count, I finally just broke down and figured undervolting the CPU would solve any (possibly thermal) problems I thought I was running into. But now all of this is coming to light, I only feel betrayed to have trusted Intel. I had a single bad experience from AMD 10+ years ago now, and it kept me going Intel ever since. But now I cannot even afford to fix this issue, or upgrade my system to AMD to fix any future issues that will occur from submitting repeat RMA's for 'new' replacements any time I get a faulty replacement. I just hope a class action lawsuit forces Intel to pay out to all those that got screwed like me (But it likely won't).
I did the same thing, ive gone through a zotac 4090-gigabyte 4090 and 1 13900 processor(all was warranty from the place minus the 100 they wanted to upgrade the zotac 4090 to the gigabyte 4090) I was getting Vram errors when I ran games and they thought it was the GPU but CPU after and havent had a problem since, I will never buy from the guys again because english is their second language and they were mixing parts I ordered for segotep chinese PSU etc I learned alot but over spent both the makers and sellers are on a hustle with PCs
Same here m8, same :( I ripped my hair off and really tried to find what caused it. Now I am so frustrated on Intel! They need to pay for all my time wasted on their fraud! NEXT AMD! NEVER Inte, they sure don´t have my trust!
I think the issue is mostly caused by shitty leadership and the policies they set, tbh. I won't even consider their stuff again til a large number of the people that are at the top right now are replaced. I usually go AMD for my own desktop computer anyway but after this, I don't see myself trusting Intel even in a laptop I buy. I'm really glad the Intel computer I built for a friend a few years back was only 10th gen.
Mate, I took a loan and bought more expensive set with 9950x because I'm going to use it for 5+ years, once I read about those microcode updates...that the hell know what they are changing inside my cpu die.
Honestly intel needs to do more. You can't expect the majority of people to even know what a bios is let alone risk bricking a pc to fix their error. It's like taking your car to get the tires changed and then the garage telling you to torque everything down to spec.
I would imagine any motherboard that's for 900K series will have BIOS flashback feature. BIOS updates aren't as scary as they used to be. I update mine every time there is a microcode or security update.
I assume people who know about this issue will ask someone to do the update for them or someone who is already tech savvy to let their family and friends know about the issue. But unfortunately not everyone will know about the issue unless they get some kind of notice.
@@msg360 two years is nothing for a CPU, especially today when the generational gains are marginal. It's not uncommon to see 5-10 year old CPUs even today, and they are performing perfectly fine with the latest software
Even if the new microcode does fix the problem, it does nothing to fix Intel's reputation. They knew about this for over a year, did not communicate this to customers, and are now apparently rejecting RMAs of CPUs bought from Amazon and Microcenter because they're "not genuine".
People today have short memories. Especially since AMD disappointed with their recent releases. This whole incident will be forgetting two months from now.
As someone who originally planned on getting a 13900k second hand to replace my 12600k in about 1-2 years seeing all of this i hope that 15 gen is fixed and on 1700 at this point 😅
I just updated to 0x125 yesterday and got a roughly 5% performance _increase_ (in benchmarks) and slight temperature drop on both my 13600K (Asus B760) and 13700K (Asus Z790). At full bore, both barely kiss 1.3V now. Of important note: I undervolt both systems as much as possible before performance loss, no overclock. The new microcode and BIOS seems to respect my undervolt more strictly than before. I'm waiting until 0x129 (or later) has a full release. Since my CPUs aren't having problems, I don't want to update them with beta microcode.
Mine i9 13900kf is barely doing 1.250v with oc to 5gzh and pulling around 195w with undervolting offset at 0.115v without any bios updates since I got it over a year ago
@@raiku9161 Go in to the bios>advanced cpu settings>set the power limit to 256w>then go to core voltage set as adaptive>set offset mode to - Then try the offset -0.050 if is stable try go lower like -0.075 or even -0.100
One comment: 13900's and 14900's have a wide range of VID settings depending on the quality of the silicon. The performance impacts are expected to be much more significant for the chips at the top end of the scale of 'how much voltage they need' than the ones that win the silicon lottery. The better cpus might not get capped at all since they are down 0.1v in a safe zone anyway. Might want to test this with a few more CPUs unless you know this one to be on the high VID side.
The Intel update specifically limits the voltage to 1.55. There have been testers who know that hwinfo CANNOT report these voltage spikes, like the hardware overclock guy, who uses oscilloscope PHYSICAL VOLTAGE MONITORS, which is the only way to see micro spike voltage. Which was going over 1.55 before patch. You cannot save degraded chips, they are permanently degraded, same with silicon lottery, and bad chips will NOT BOOST to the max especially with voltage limits.
It wasnt released overnight. Intel knew about it for a long time and probably knew how to fix it and how much not to push. Besides that we still dont know if this will fix it.
@@coolvinayExactly, no one can tell for sure, the main point of this patch was to make the CPU's last long enough, could very well fail in the span of the next 4 years; for example these generations won't retain second hand value and it is debatable if Intel would keep the suggested prices or cut them down. It is already a worse deal based on the amount of uncertainties.
relevant take, im guessing microcode revision targetting specific variables without losing performance or functionality is not just a trip to jiffy-lube lol.
I've had zero issues with my 16 month old 13900 non k. It's about 5% slower than a 13900k, but it's been rock solid and stable. I think the slightly lower power profile has kept it relatively safe. I'll probably still get the micro code eventually, but I'm not too worried, not to mention Intel has extended the warranty for my CPU until April of 2028. I'll probably be upgrading my CPU in 2 years anyway, but a nice plus.
for someone who's i9-14900KS that's been working just fine, I owe them a Thank you for putting out a fix that will PREVENT said CPU from crashing. You sound bitter asf.
I have an Asus z790 dark hero with an i9 14-900K. It is brand new (June 2024) and i used the 1302 bios version since day one. Everything was working extremely well. My voltage was around 1.330, and my cpu never went over 70 C. This past friday i decided to update to the new bios (just for the new microcode). It is not good. For the first time since i built my pc, the cpu reached 80 C. I even got an alert from my AIO. The voltage change to 1.455. I use it mostly for gaming, so when i tried to play some Warzone, my cpu was getting extremely hot. Before the update the cpu stood around 65% to 70% TOP. But now, for some reason it was reaching 75 to even 82 C. That's when i got this alert from the AIO telling me my cpu was too hot. So i decided to just go back to the 1302 version, and now it is back to normal. Sorry for the long post and my bad english, just wanted to share some info about this.
I just did my 14700k which was running 5ghz P cores and 4ghz E cores, 1.12v and 175w during CB r23 before the 129 microcode. Updated to microcode and my 1.12v jumped up to 1.35-1.42c and 215w during CB r23 with a 2k performance loss. Went back to the bios before and still running super cool and better volts and watts and regained my 2k in my CB r23 score. Either I got a golden sample or running it this way is the best solution to those at intel that think they need to push clocks so high to compete with AMD that they forget about longevity of the average to hardcore PC user. Before this I was using an 8700k so keeping the P cores at 5ghz and E cores at 4ghz is a massive step up for me. These CPUS never should've been marketed at 5.7ghz to 6ghz on the performance side as intel chip isn't as good as TSMC who makes AMD's. I'll stick to my setup and not worry about the issues for running it too fast, too hot and getting degraded over time.
Small side note here. In physics Power is always produced by multiplying voltage times current or amps (ampere as the french say). So, we have V * A = W(atts). That means, there higher the voltage for a current is, the more watts you have to cool. Before I RMA'd my 13700K, I was watching Buildziod's voltage analyses video and I thought to myself, what if I fiddle with the resistance and voltage directly in my VR-settings? So I did. I set a upper voltage limit to the VR (voltage regulator, which you can fiddle with since Z690) to 1.380V, which means, my 13700K got an upper voltage limit to 1.380V to run at stock speed. I also lowered the resistance settings in the VR-settings from 40 / 90 AC to DC loadline to 10 / 10 AC to DC loadline. And I set the current limit to Intel's specification to 307 Amps. With benchmate I got my 13700K from 29600-ish point in Cinebench23 multicore point to 30200-ish points at stock and 30962 point with a +300MHz overclock and +200MHz on the P-cores and E-cores, but I had to adjust the voltage limit up to 1.450V. However, within Intel's spcification and my self-emposed voltage limit, Cinebench23 did'nt even get remotely close to the power limit of 253 Watts. Highest measured was 233 Watts and the highest core temp was at 82°C. That is a world away from the previous 104°C.
I'm running my 13700K paired with a Z790 MSI Gaming Plus WiFi motherboard and a DeepCool AK620 air cooler, at stock ratios, 110/110 AC DC load lines, ICCMax at 307A, IA CEP=enabled, and a global negative voltage offset of -150mV. This applies to both the CPU core and Cache, lowering their VIDs. PL2 is capped at 188W, max temp in CB R23 reaches 85C with 23C ambient, R23 score is 30200 pts, the highest I've seen is 30313. VCore during this multi-core 188W R23 load is around 1.18V. If I set PL2 to 253W, the max draw is 215W and the score is around 31K. Microcode is 0x123, currently running the latest stable version BIOS for my MB. Max VCore spike I've ever registered (HWInfo is running 24/7) is 1.338V, during light low-core usage. People online tend to mostly undervolt through lowering the AC load line, but my personal experience is that I get the best results with keeping AC=DC=110 (those are MSI's "Intel Default" settings), and then applying a negative offset. No instablity at all, I've successfully passed all kinds of tests such as y-cruncher, OCCT, Prime95, Cinebench R23 and 2024, TimeSpy Extreme, Intel's Diagnostic Tool, and of course everything is stable during regular usage and gaming. I can't really be more happy with my CPU and my configuration, hope it stays that way in the long run.
Purchased the 14900k in Feb 2024 from Microcenter. Was having non-stop crashing during that time trying to play Helldivers. I saw your video about the power settings and that solved the issues of crashing, but still had program stalling. Updated to the 0x125 yesterday on my MSI board. Still having issues with program hangs during high loads. Tried running Cinebench and explorer will hang and crash. It's insane. I will be heading to Microcenter to replace the processor on the weekend. I hope that fixes the issue.
If you have still having program stalls, then you're still getting computational errors from the cpu, which means your CPU is really sick and still dying even with the new power settings. You should definitely RMA it when you can.
Are you using a high power profile in bios? My default power profile is basically unlimited power on MSI board it is unstable, set to intel defaults and helldivers etc is all good.
4 threads were fried in the process of testing, turns out the degradation is going at slightly faster rate than originally anticipated. Still good for 7 more tests, remember NO REFUNDS
@@rustler08The cores that reach + 1.5 volts are the P-cores and the ring voltage is the same as the core voltage. The ring can't handle the voltage, the cores can, i learned this from leaks from Intel engineers in Moore's Law Is Dead, they said that they know since Alder Lake that the ring can't handle that much voltage. Considering that Intel wanted to launch Raptor Lake fast, that explains the quality control issues.
Just launch an Unreal Engine game while cpu is OC if "memory error out of video memory" and the game closes. It's still a cpu fault. If it BSOD its a cpu fault.
just run it stock watts for the sku you have. if it doesn't deliver the advertised turbo performance it's trash. technically or marketing error. upto no good clause.
@@glitchered That doesn't demonstrate whether any 13th/14th gen CPU has been damaged or is suffering from this issue, all that will do is demonstrate that you lost the silicon lottery. A better test is repeatedly performing any decompression task larger than 300MB 5-10 times. If it errors out, or the time remaining goes into hours or days, you have an affected CPU.
I bought my first Mac. Macbook air M3 Base, only 8GB ram and i am choosing to complete most tasks on it. I cant believe how much better everything is on this thing.
I would honestly take a 5% performance hit if that means there really is a significantly lower chance my CPU will burn itself down. But yea, even if intel announce a long-term fix, we wouldn't really know whether it works until maybe months or years later.
The CPU’s are dying from Intels dumb stock bios setting which pushes 1.5 volts to two of its cores. If you change its stick settings it will never degrade.
On stock settings my cpu was pushing 1.495 v on the vcore With the 0x125 microcode update and a slight 300 mhz underclock paired with better ac/dc loadline settings it never gets beyond 1.14v and is very stable They dropped the ball with their defaults at launch
they should offer EVERYONE a new chip to replace it with the old one. & then u can sent it back after you recieved a replacement. 1st Faulty product or working product that was ran out of spec by intel. 2nd false marketing. 3rd its already semi broken for some people since the chip degraded & it will not guarantee a long life expectation. a new chip will may last 6-10 years while the old chip can break down in the next 1-3 years.
@@Grievous- Cool is a relative term. A 95 degree day is cooler than a 100 degree day. See also: "Cheap mansion", "cheap supercar", etc. Let me know if you're still (somehow???) confused.
aaaaaaand since this "fix" is going to require months of failure data to see whether it's had any impact on reducing the rate of degradation - ONLY FROM THIS POINT IN TIME ONWARDS - this issue is guaranteed to continue being unresolved and a total PR fustercluck for Intel for months to come
Never been so happy to have ditched intel. Degradation won't continue at an accelerated rate for the affected CPUs, but shortened lifespan is a guarantee.
quote: "...because you are not using a genuine one, you are using fake cpu...." as if it was so simple as manufacturing a pair of shoes...Someone went too far with its statements. Too far.
I undervolt my CPU as many people watching this video needed to do since the heat the 13th and 14th gen CPUs can generate otherwise. My undervolt settings do not seem to work as well as they did before. I was happily undervolting by an offset of -0.070v prior to the new microcode. It was just a simple and modest undervolt by setting the negative offset in the ASUS bios and could run the Cinebench 30minute stability benchmark just fine. It worked for well over a year. Now I can only apply a -0.050 offset. It does seem to still keep the temps down but I'd love some additional footroom just in case. Cinebench doesnt go above 95c in the 30 minute test which is similar to before. Maybe its just that the update addressed some of the exxcessive voltage and I no longer need as high of an undervolt?
Very interesting Data. Just goes to show, the high voltages weren't at peak power loads, and not peak workloads (ie. multicore). And with the minimal performance drop with the new patch, you can really see that these high voltage spikes weren't really necessary. They didnt' give you more cinebench performance, they gave you "better performance while loading" ie, when no one really cares or notices. These voltages were to hit clock speeds that were largely marketing bulletpoints, which minimal performance benefits. However what they did do is degrade the silicon. Hopefully Intel learns their lesson.
Yeah they learned they need to slow the planned degredation of the next generation by just a few percent so the chips don't fail until after the end of warranty.
With a very old BIOS my 13900KS never exceeds 1.4 Volt. So, I'm definitely not affected. The CPU runs heavy multi threaded and also heavy single threaded workloads at a daily basis with no issues for 17 months now.
@JayzTwoCents MSI just dropped another updated BIOS today (8/13/2024). You may want to retest with the newer 0x1 129 BIOS. I retested my 13700K on an MSI Z790 TOMAHAWK WiFi6 DDR5, and I saw a significant boost in the Cinebench R23 multicore score. MY BIOS settings are set to Intel Default Safe, XMP1 bringing DDR5 to 6000mhz, and CPU capped at 253w.
This isn't like a GPU where you could go to your closet, pull out an R9-390, and have a usable (but hot and slow) experience while you wait for the RMA+shipping both ways. A person who has the means will go out and get an AMD MB with a 9700 or 7800x3d and do a system transplant. *They might not go back to intel ever.* The 'old' intel MB and when (if) they get the CPU will end up as someone's "gift"
considering amd is running into issues on their new chips this is nothing new to any company, both amd and intel, nvidia and so on have all ran into multiple issues
@@yulfine1688 i am pretty sure many people choose Intel because of their reliability, plug-n-play superiority compared to the mess from Ryzen memory bs, especially their 1st gen. guess now Intel is expensive, hot-power hungry, and unreliable. surely a winning combo for any product LOL
@@MuhammadRidwan-pe7ny intel has always been on the power hungry side, heat has been a mixed bag, for the rest that hasn't been the case for almost a decade now, amd has done a generally good job. Also they have been mostly neck and neck with intel at times having a slight edge until the x3d chips which sometimes have an edge. People stick with intel because they're large and well known its a comfort pick. The average person is also usually stuck with intel because most laptops and so on are intel, though we are seeing more amd laptops.
meh. Y'all acting like AMD has never had its fair share of f ups. Funny how all of those are magically forgotten about right now. All Intel has to do is release a good chip and all of this will be forgotten except by terminally online people who tie their identity to a corporation. Everyone is jerkin AMD right now, but they really did used to be an awful brand. They made good chips and now everyone likes them. The idea that Intel can't recover their reputation is just ridiculous
Alright 16:03 how long did it take you to get that shot lol. If that was cut from two pieces of footage that's one of the smoothest cuts I've ever seen H O L Y
Intel today is not the company that pulled their halo pentium chips for the famous alu bug. The company had better leaders then who wanted to please their customers not the stockholders. No corporation deserves our loyalty, plain and simple. Corporations today are money making machines for the elite stock holder not you the customer. Appearances and pr is more important today than a lasting product. Blame it on ideas like manufactured obsolencence and closed systems with no right to repair. I would apply the same for AMD also. Brand loyalty is not a good long term strategy for the consumer.
I’m a stock holder. I don’t have enough share for the dividend to be shit but I would still prefer good products. I invested in Intel because I use their products. Bought some more of this cheap stock too. I don’t recommend that
@@thepopeofkeke Obviously the typical individual stockholder has not much voting rights to influence. I was talking about the institutional investor, the Blackrocks and Vanguards of the world and the elite investors they represent.
@@danielgarciasalorio8646 I cant speak to the new platforms but for am4 gigabyte is actually on top of the game for motherboard updates. My first gen am4 board still gets regular updates and normally before anyone else. Its possible that they are very different teams though haha.
MSI has been trying to step up their game lately. I wish motherboard manufacturers would start competing again, stagnation is real, basic features are being sold at a premium these days and it sucks
Im on old microcode and no plans for updating bios ... Tuned performance and thermals myself ... ACDC Loadline 28/55 LLC Medium , core offset -0.055 V , 253W PL1/PL2 ,CB23 score muti 39840k on gibabyte Z790 gaming X AX @ 13900K , no thermal throttle , stable as hell passing Ycrucher ,and CB 10 minutes and 30 minutes runs ...VID MAX never exceed 1.360V , VCORE MAX pek 1.380 .Amamzing performance and low temperatures !
Yo, Jay you are 1 of the few Pc UA-camrs I actually trust. I trust your judgment I know sometimes people can be wrong. But when it comes to you man, I know that you are here for us, you try to do what’s right as far as information goes. Thank you for helping us!
Jay is fun to watch for fun projects but for straight up fast solutions, I would watch FRAME CHASERS. FC's found and fixed this whole Intel issue last year. It's just too bad he didn't have a massive following when he had the Intel fix.
Remember, we just want strong Intel and AMD, and if it were possible, others, competition is key. Yeah so I think it’s reasonable that people expected a drop in performance, without being sure it would.
@@conza1989 The 13900K multicore is faster than a 32 big core Threadripper 2990WX, i don't know why people would complain about a 7% drop in perfomance with a 100 watt reduction in power consumption... That is still, way, WAY too fast for a consumer CPU, unlike Threadripper, extremely fast for games as well.
I advise people to limit the power in bios and wait till more testing is done, with these rushed stuff you never know they will not pass throught the usual QA and will end up fucking up.
It's weird, because actually, for my 13900K it was more stable without limits and overclocked than with a limit, at least while running Cinebench R32. I tried the limits when people started saying to put the limits, and it was bad on Cinebench and had crashed after a few runs. Never had a problem with games or apps just while torturing the CPU on Cinebench. Without limits runs great like it has since I bought it over a year ago. But after I tried the microcode 0x125 a few weeks ago and its stable no matter what I try! I get a score of 40200 in Cinebench R32 without limits and I can run it for over an hour without issues! so, I am not worried. and probably won't even try the new microcode since it's so stable now since the microcode two weeks ago. Anyway, planning to change it to a 14900KS soon.
Just limiting the power will most likely do jack S#!T, since the main problem is the high voltage. Which is present when a single core boost happens. Which, guess what, is when the processor runs at 30-50W. When at 200-300W, the voltage is actually lower, just like the frequencies are lower at all-core workloads vs single core workloads which can boost to 6.0 GHz.
A lot of the lower quality high end Raptor Lake CPUs can request more than 100mv extra, compared to a good sample, and those will be the ones that see the most degradation. Better samples will have lower VID requests, so the even the worst spikes on the old microcode might not even damage a good sample. The question is what voltage bins had VIDs high enough to be vulnerable? The worst 10%, 20%, 50%?
Seems mostly margin of error stuff. The real question which we'll take time to get the answer for is whether this will actually fix the issues or not. Jay also only really tested the 14900k. Performance deltas could be different on the other stuff.
They trying to fix it for last two years... so this new "fix" released in the same time when AMD showed their new stuff looks to be PR fix not hardware fix...
Yes. To give them the benefit of the doubt, I sent a support request to MSI for a beta bios (in their forums they were saying you have to ask for it) but they answered, "sorry, no bios for the motherboard we sold you." This represents my first and last customer service experience with MSI. Live and learn.
I have a 14900k on Asus z690-E and I've had it blue screen, game crash and chrome crashes in the last few months. I will wait for x129 update to come out officially and see if it makes a difference for stability, otherwise if my CPU is degraded I will go through the process of RMA so I cant wait for that headache.
GL, If it has degraded that would be the best. Not the same issue but my 12700KF & 12900KFs got stable e cores off, turbo boost off (causes some games to crash), switched to Linux (been gaming on Nobara past few months) Asus Z690 TG Wifi. Same machine(s) would BSOD previously just the 12700KF months ago in Firefox not doing much it was the last straw for me to try more things. If degradation has occurred would RMA though.
If running a Windows Preview version, it might be a good idea to unable encryption/de-crypt drives before updating UEFI. Bitlocker might send you on a recovery prosess for unlocking your drives. Might be relevant as bitlocker "light" gets enabled on Home versions.
Worst for the voltage control are quickly switching workloads. All-core load are not most stressing for modern CPUs (i think since Zen3/Comet Lake), because they just won't push clocks as high as in single-core loads, so they don't need as high voltage, even if they consume most power they can. There are tools like CoreCycler to test switching workload behaviors.
@@martinxyz Not even close to the first one to notice it. Every medium-large game dev knew because of the sheer volume of crash reports. Ive been having issues with my 13900K since the start, and I ran into specific pages for 13th gen CPU related crashes on every game I tried to resolve them in. Quickly became clear it was my CPU that was the issue. We've been talking about it for well over a year, but happy to see it finally get into the mainstream.
@JayzTwoCents 8:28 Please move the lower end of such Graphs up a bit in the future. If i pause the Video, the labels are just under the youtube video controls that are visible all the time when paused.
I work in IT - No offense but given Intel's frankly shameful response on this issue I won't be "canary" for it's microcode update. Given it's also flagged by MSI as "beta" I'll wait until they've *fully* tested it.
Im running a 13600k. I recently updated to the new microcode and i started getting higher temps in benchmarks. I have seen a couple people report that MSI has boosted the lite load in the bios to still draw more power. I verified this myself and saw lite load was at 18 when 12 was the old default. Turning this down can greatly improve temps without hurting performance.
It would be super helpful, if you'd be willing, to put out a video for those like myself who are now having to learn how to update bios and adjust settings in bios because of the Intel blunder, showing us how to do so safely. I know with ASUS boards you can use AI suite 3 but some people say I shouldn't. I'm having a hard time learning properly with so many contradicting opinions, it'd be nice to see some of the bigger tech youtubers like yourself put out videos to help those of us who are much less read on this subject.
If you have an MSI, don’t bother with the update. The update essentially solves nothing. Permanent processor damage has already been done. Contact MSI about complete system replacement
I am waiting for the dust to settle before I download anything. No more beta testing for me. I have other computers until this is resolved. Luckily all I have done is install the os in the msi computer. So hopefully it hasn't been damaged. The other one is a gigabyte and I haven't bought the processor yet. May wait a year at the rate they are going.
Got to give props to Intel on the RMA process, no questions asked after they read instability issues. They also overnighted the product to me once they received it.
Did the latest update on a friends 0x1 129 for I9- 14900K CPU using an nvidia 3060 video card and couldn't even get into the bios after the update. He didnt want to try NVIDIA GPU Firmware Update Tool for DisplayID so Installed the bios update from May, prior to all the intel updates and now it runs fine.
Asus states in the BIOS descrption "The new BIOS includes Intel microcode 0x129 and adjusts the factory default settings for the non-K processors, enhancing the stability of Intel Core 13th and 14th gen desktop processors". So, this would be useless for anyone with a K or up processor?
Hi, So from my testing, my CPU was always crashing randomly with the default settings ( I have 14900k ). He said that if your CPU was crashing then its already dead. I don't think this is true at least at first glance. Now I do not know if the CPU will keep degrading, but I can tell you that after the 0x129 microcode update of my BIOS which was 7 days ago, my CPU has not crashed once while gaming using the default settings. I have not set it below 1400V ( which I had to do before to be able to achieve stability ). So based on my experience I think that the microcode works just fine for K processors ( Although I have a Gigabyte motherboard ). What I noticed as a minus for me at least is that with the default settings the CPU is a bit hotter while gaming, but at the cost of some performance if I reduce the voltage the temperature will drop as well. Will not touch anything for now though, want to keep gaming like this and verify if I will have any crashes. Of course now we just have to wait, I guess time will tell if it is really effective, but it definitely looks better for me at least for now.
"The latest microcode update (0x129) will limit voltage requests above 1.55V as a preventative mitigation for processors not experiencing instability symptoms. This latest microcode update will primarily improve operating conditions for K/KF/KS processors." From Intel page
My 13600KF CPU is damaged, and I just sent it in for RMA yesterday. Every Cinebench test resulted in overheating, reaching 100°C, even tested a few different cooler and the latest BIOS from ASUS, default intel setting. Blue screen randomly in a clean window. So much for the legendary “Intel stability,” hahaha
that s poor on your side bro; either the cooling; or you still changed something in bios settings; i have 80 degrees on my 13700kf; be sure that ur bios is on default; and you can set anytime max cpu temp in bios; unless u bought a shitty motherboard and now wonder why ur cpu is getting cooked
Don't use default Intel settings with 13600kf, those just result in over voltage and overheating. Same with my 13600kf, use the motherboard defaults and undervolt it if you can, way better
Fixed my MSI / 13900k setup. I had to press and hold the power button down a few times after the bios update. I also applied the FW update under the bios. Once I was back up I had to reset my Microsoft Login info. All good now! Don't forget to also go back into Bios and turn on XMP.
Not to be intels fan boy, but i hear alot of "bad stories" by youtubers and read on different forums about INTEL's RMA process. I just got a brand new CPU sent from Intel, i wrote their support saying that i was worriesd about degredation, they told me to send it back and they would send a new one to me. The email's back and fourth took 2 days, Intel paid for a delivery service to come and collect the old cpu at my address, they paid shipping, and the cpu had to shipped to a different country. The cpu left my house monday a 11, and a brand new one arrived thursday around the same. Just saying, if u got any worries about your cpu, dont hesitate to write to their support, this is the most painless dealing ive ever had with any company about any returns, super nice and polite support team, and super quick as a i was expecting to be with a cpu for atleast 14 days. Side note, thx for the video, informative as always. Hoping u gonna do some OC testing on an asus mb sometime in the future with this new microcode.
Seems like it goes both ways, I read a comment from someone who is having a terrible time getting his CPU RMA'd, they wanted pictures of the CPU, attempts to reset the bios, and some other junk I don't remember and they still haven't approved an RMA for a 14900k. He said it felt like they were purposely delaying him. Does your country have strong consumer protection laws?
@@volvo09 I live in EU, Denmark to be exact, i couldnt say if we have any perticulary strict laws in place, compared to other countries to be honest. Also i dont know if they apply here aswell, as im thinking it just falls under to the standard 2 year warranty that comes with the product. B ut again, im no expert here, so couldnt say if where i live is a factor. Sad for that other guy though, truely the very opposite of what i encountered.
People absolutely HAMMERED AMD for doing that around the times of Bulldozer. I do find it quite ironic the way that the reverse is happening currently, and the people who were vocal then, are quiet now. I've never advocated fanboy behaviour, but people should at the very least be consistent.
LOL, no. I was one of the people dunking on AMD back then (bulldozer and especially PILEDRIVER) for 'HURRDURR MORE CORES', and now I'm dunking on intel for selling defective-ass CPUs with horrendous power inefficiency and trying to brute-force the problem. There's plenty of consistency, you're just not seeing it.
@@NapalmFlame You're a minority, I think. For some reason, in a similar manner to the brand loyalty people provide in the GPU market, it seems that people feel obliged to "Pick a Team" I can't say I understand it, but it is what it is and I do think that there is a large majority that do so.
I have same mobo and just updated to test it and what i noticed that there isnt anything new i tweaked most of those things manually and actually i got 35k cine score with new bios stock power and 82C on cinebench before i had all cores lcoked 5600 and 4400 with manually tweaked voltages and i got 71C and 39700 cinebench score lol
MSI MPG B760I Edge Wifi - updated my bios due to the Intel Microcoding update. After flash system booted fine. All CPU settings and ram settings correct. Later in the day after work loaded up some games and no sound. Reinstalled all audio drivers and still nothing. flashed back the bios to 7D40vA8 and still no audio detected by Microsoft. Even reinstalled all Motherboard drivers, windows 11 up to date with supports. Created a ticket with MSI but I fear Audio Chip is FRIED! Keep up the good work Jay, you did warn us to hold off on updating lol
@@djcetra will you notice if its money? A 14900k is like $550 and a 14700k is around $380, so you lose $170 worth of performance in one update, that sucks. 🤣
Yeah AMD 9700X is way better overclocked than stock. See Der Bauer video. 5400mhz all core and it will pass Intel. So IPC is a tad better. It did CB23 -> 24000 points, my 13700K with E cores disabled did around 23000 points.
Jay finally delivering the content that community needs and is actually the only creator that i found that seems to do it right and early. Big props! If there will be further microcode updates for this issue, see if you can find even more scenarios on how to test/replicate the 1.6v bursts and new behavior.
"We have sold you a car with the front wheel axle sawed in half. You can lose steering and crash. We have found a brilliant solution: a firmware update that limits the speed if there are holes in the road or if the shock absorbers shake too much, also the speed in general. You can still end up going down a cliff and die.... in that case we'll see what to do on a case by case basis if we feel like it. Don't complain or we won't help you anymore." This is great reliability and customer service. Who knows why they lost 40% of their stock value in 1 month and 50% in 6 months?
I run my 10900K exactly the way Intel is saying is not safe, no power limits, no c states. Still rock solid, same OC as day 1. Intel has totally ruined their reputation over this scandal. I've been an Intel Andy because of their reliability, compatibility (no weird XMP issues or CCX concerns), and stability. Now I am not sure exactly what my next CPU will be.
I know everyone's pissed at Intel, but if they can fix this it is excellent news, and we should all be rooting for them to rebound. If Intel fails and AMD has no competition, CPUs will become extremely expensive, and progress will slow significantly.
i mean; 5600x was 30% faster than 3600x; 7600x was 35% faster than 5600x; 9600x was... 1% faster than 7600x ? Thing is intel was always good on platform jumps; which should be the case for arrow lake too; then they were always like filler gens or refreshes after the new platform; maybe this will change
Already updated on the Gigabyte AORUS board for myself. Glad to know I am on the safe side at least for now. Have not had any crashed or instability on my 14900K.
Thanks for the video. Would have liked to see a frequency comparison for the Single-Core stress test workload, as I believe you only provided the 10 minute all-core workload chart. As you stated, the single threaded is where you would notice differences in frequency the most (and see if 1-2 cores can still even hit 6ghz anymore), and would have a significant impact on gpu-limited games.
@@EightPawsProductionsHD No, because people found that with the 11900K they got what this guy has.... An 11700K, and got pissed off... Dare we say when AMD was shitting the bed with $100 2600x and $125 2700x, AMD boys was telling people on reddit to buy 10400's and B560 motherboards. Then the same people saying that crap was on Discord building their own 10400 B560 rig LIVEstream and finding out their top M.2 wasn't being used... After telling 1000 or more people to buy that same setup....... It says so on the b560 specs and in the manual.....
*_"Just sitting here sipping tea with my 11700k with zero problems and no “efficiency” cores"_* Just sitting here smiling with my 13700k with zero problems, because I'm smart IT guy, and had it slightly underclocked from the start, instead of allowing the motherboard manufacturer trying to blow it up with its settings in an attempt to score great in reviews. It scores 46,494 in multi cores benchmark vs 23,963 for your 11700k. It scores 4,362 in single core vs 3,406 for the 11700k. Google: "Puget says its Intel CPU failure rate is lower than AMD Ryzen failures" *Chart* AMD Ryzen 5000: around 4% failure rate AMD Ryzen 7000: around 4% failure rate Intel Core 10th gen: around 0.5% failure rate Intel Core 11th gen: above 7% failure rate
Test should also include the version with uncapped mobo settings, because that was the default running mode for most people pre shit hitting the fan. There were a lot more noticeable multi-core performance drops when switching form uncapped to Intel Extreme Profile, which lowered power consumption about 60-70W average (sometimes even more).
The amount of power the world is wasting by running Intel atm is crazy. I get that people don't care in areas where energy prices are low but damn, those suckers produce some heat.
My 14900k using 13w at desktop 70w in games, my 4090 on the other hand is what heats up my house 😂 cpu power draw really a drop in the ocean compared to GPU
I have a 13400F that hasn't had issues so far. It's also true that we're talking about a rebranded and locked last gen 12600KF, so it shouldn't be affected in the first place
I just did what Buildzoid did and limited my CPU Vcore input max to 1.4 volts (plus a couple other LLC tweaks) and everything's been stable since. Granted, it doesn't boost over 5.7GHz anymore, but I'm still getting mostly the same Cinebench results as I was with the unlimited/4096A power limit settings, so I'm fine with not getting 6GHz boost if the overall performance is the same. I'll still check out the 129 microcode though when ASUS releases it for my board
@@haies09 I tried that on my Asus ROG Strix Z790-A board but it completely killed performance. Like I was only getting 10% of what I should've been getting. I think that setting is specific for Gigabyte boards
Is there a tool to determine that there is or is not damage? The Intel Processor Diagnostic Tool does not clearly show that it is or is not checking this.
How do you want to check for a failure that so far has had about 4-5 different reasons. Run some performance intensive workloads or run some benchmarks and see whether it crashes or not.
As far as I know, If you run a benchmark and look to the frequencies and voltages of each core (you can use hwinfo or cpuZ). If all of them are in the expected range for performance or efficiency cores, them there is no damage. However If one is getting hotter than others or If your system crashes onde in a while, them you have a problem
nope, unless u experience crashes or bluescreens with stock settings. it may already degraded - no chip is damaged - they all degraded to a point it gives out errors & then ppl call it "damaged chips"
I have a bit of professional experience in semiconductor manufacturing and reliability/damage. The physical damage is going to present as erosion (causing leakage) and / or shorting between transistors and traces. There are systematic electric testing that is done in QA prior to packaging that can find such problems, but I don't know if these tests can be replicated purely in software post-packaging. Since damage is not binary, the quantum well errors must be statistically calculated over a fairly large number of test runs. With 4.2B transistors, this would be a long and involved test, but again, it's unclear to me if exercising specific cores and transitions is possible without physical access to the test pads on the die. If any semiconductor test and QA engineers are reading this, I'd love to get your opinion.
Dude, thanks so much... I only have the 0x125 BIOS update. I updated the BIOS, but I never updated the MSI ME firmware...... I noticed the small font and downloads, but I didn't do that until I heard it in your video... 0x129 isn't available yet on my Mobo... thanks man!! I'll keep checking for the newest.....I'm running a 14700K... never experienced any issues... but it's a fresh build, less than 3 weeks old...
I'm worried that my chip is already fucked. I know they extended the warranty to 5 yrs basically, but I had ghost of tsushima cause a bsod once (only game or app that has done this) Worried it's already fucked and have no idea how to tell if it's actually messed up....
The thing about this and what most UA-camrs have touched up on is that any possible damage is irreversible. BUT if there aren’t any more issues since the BSODs you mentioned, then there shouldn’t anything to worry about since there shouldn’t be anymore degrading
Basically uninstall your gpu drivers and reinstall them 10 times, if nothing crashes it's probably good. Zipping/unzipping are workloads that are very sensitive to this + more than often decompressed files are checked against checksums
i5s already had lower voltages either way; bcs they re boosting way less than an i9; kinda same as i5s for the i7s; that s why they were at a lower risk
Yeah was wondering about the 13600k too, despite it being safer. Bought one recently, and I set a negative offset for voltage soon after ('cause summer ambient temps). Apart from a not so great single core score on cinebench, it's been working great, so I'm in no hurry to update. I'm sure someone will test the new microcode on them soon enough though.
A patch for a chip that is already permanently damaged, and also effectively beyond repair. Intel is going to be writing volumes of refund and replace checks. Pre-built manufacturers such as Dell will also be doing the same. At a minimum imagine what those checks will be like for: the minimum of a replacement cpu and motherboard. If it’s an item like a pre build (given how many items are permenantly attached) it will essentially mean replacing the entire system
Thank God you post some videos about motherboard manufacturer are making processor broken. I just recently bought i7-14700k and learned to undervolt and use the Intel default profile.
The proof that Intel issues are manufacturing defects is the 13th gen cpus crashing now. It seems those CPUs are aging due to issues in the manufacturing process. Let's see. I hope all issues are fixed and Intel learns to change
I'm just happy to see these finally coming out.
On a side note AI summary is getting wild.
"Intel released a microcode update (0x1 129) for the 14900K CPU to address voltage stability issues. The update resulted in a slight performance drop in some tests, such as Time Spy Extreme (700 points), but no significant changes in most other benchmarks. The update also reduced voltage spikes during single-core workloads, potentially improving stability. However, the author notes that it's too early to say if the update fully resolves the voltage issues and recommends waiting for further testing."
Wow which AI tool summarized the video? It sure did a better job than many people watching it since it doesn't apply any bias to its view!
@@Jayztwocentsit's the " Ask" button that shows up under the video in the default Android UA-cam app on my pixel 8 pro. Basically auto summarizes most videos. It doesn't show up on all videos, however I notice it now far more than I used to. Nearly all at this point.
So - to more directly answer it's probably Google Bard in some form
How’d you obtain this? Via an api? What’s the api called?
@@kanyeeastlolzPunkAssUsesAIv2
They should also release a testing executable to perform stability tests. By the amount of data they have at this point, they should know what to look for.
they would prefer that you forget about this now
Intel PDT (Processor Diagnostic Tool) already exists, they'd just have to update it with a test that explicitly produces errors if your CPU is affected.
@@styleisaweapon Nothing to see here, move along!
@@styleisaweapon Lol i was thinking the same thing, they tried to reject RMA as much as they could, so why would they release a tool to identify stability problem ? That would only increase the amount of people asking for RMA 😂
Intel already has for a few years, its called Intel Processor Diagnostic Tool one can find it easily.
>It's time to RMA that, however painful that process is gonna be
I'm in the middle of a RMA for my 14900K and Intel loves nothing more than employing dilatory tactics. It’s been weeks of drawn out emails, taking photos of the cpu, sending invoices and receipts, and they’ll do anything other than help me. I would have gone AM5 had I known how badly this was going to turn out.
Class action suits are being filed, hop on the train
@@jamesrose96 That's what I'll be doing
@@jamesrose96 how does one "hop on the train?"
this is why i dont buy anything for the warranty. every company does all they can to not pay out so why bother with any amount not worth getting a lawyer involved on.
@bradhaines3142 warranty is another word for insurance. So replace warranty with insurance and u will understand why the rma process is so bad
My only question is if they knew this was a thing for basically 2 years why have we only just now got a microcode update
Because randomly changing the microcode early would have brought unwanted attention to the issue, they were hoping to keep it swept under the rug.
Because Intel is a scummy company, theyve proven this countless of times.
Amd had an issue with voltage and caused CPU’s to literally catch fires it was patched within days IIRC. 7800x3d all the way.
Because the common user was getting to aware of it.
It could no longer be ignored.
Class action lawsuit
I was one of the early adopters to 13th gen intel. I made the mistake of going cutting edge while Covid was keeping everyone home, so I blew over 4k on a new 13900k and 4090 rig I built myself. I ran so many tests and diagnostics, tested thermals, edited windows configs, REINSTALLED windows like 4 times, and reflashed bios. I reinstalled drivers and even replaced my RAM and motherboard entirely. But no matter what I did, I kept hitting bluescreens and MANY game crashes that I just chalked up to be faulty 4090 graphics drivers that have yet to be patched into a playable state. Only just now, MANY months of pain and anger, am I finally finding out what happened to my build, and why everything went sour.
Intel's quality control, and now PR trying to cover up the issues for the last ~5 months, has indefinitely lost them a customer from me. I will never be supporting Intel in their shady ass business practices again. Now that its been over a year of having this PC and I've ran into too many headaches to count, I finally just broke down and figured undervolting the CPU would solve any (possibly thermal) problems I thought I was running into. But now all of this is coming to light, I only feel betrayed to have trusted Intel.
I had a single bad experience from AMD 10+ years ago now, and it kept me going Intel ever since. But now I cannot even afford to fix this issue, or upgrade my system to AMD to fix any future issues that will occur from submitting repeat RMA's for 'new' replacements any time I get a faulty replacement. I just hope a class action lawsuit forces Intel to pay out to all those that got screwed like me (But it likely won't).
I did the same thing, ive gone through a zotac 4090-gigabyte 4090 and 1 13900 processor(all was warranty from the place minus the 100 they wanted to upgrade the zotac 4090 to the gigabyte 4090) I was getting Vram errors when I ran games and they thought it was the GPU but CPU after and havent had a problem since, I will never buy from the guys again because english is their second language and they were mixing parts I ordered for segotep chinese PSU etc I learned alot but over spent both the makers and sellers are on a hustle with PCs
Same here m8, same :( I ripped my hair off and really tried to find what caused it. Now I am so frustrated on Intel!
They need to pay for all my time wasted on their fraud!
NEXT AMD! NEVER Inte, they sure don´t have my trust!
I think the issue is mostly caused by shitty leadership and the policies they set, tbh. I won't even consider their stuff again til a large number of the people that are at the top right now are replaced. I usually go AMD for my own desktop computer anyway but after this, I don't see myself trusting Intel even in a laptop I buy. I'm really glad the Intel computer I built for a friend a few years back was only 10th gen.
Mate, I took a loan and bought more expensive set with 9950x because I'm going to use it for 5+ years, once I read about those microcode updates...that the hell know what they are changing inside my cpu die.
Honestly intel needs to do more. You can't expect the majority of people to even know what a bios is let alone risk bricking a pc to fix their error. It's like taking your car to get the tires changed and then the garage telling you to torque everything down to spec.
Extended warranty for those effected Bios fix for those who have working CPUs what do you want them to do?
I would imagine any motherboard that's for 900K series will have BIOS flashback feature. BIOS updates aren't as scary as they used to be. I update mine every time there is a microcode or security update.
@@aetherfox4404 yes, my MSI motherboard is super easy but my brother running asrock bricked his pc...depends
I assume people who know about this issue will ask someone to do the update for them or someone who is already tech savvy to let their family and friends know about the issue. But unfortunately not everyone will know about the issue unless they get some kind of notice.
@@msg360 two years is nothing for a CPU, especially today when the generational gains are marginal. It's not uncommon to see 5-10 year old CPUs even today, and they are performing perfectly fine with the latest software
16:08 Phil was so proud of this shot
All hail Adobe ;)
That was an awesome shot
"yo your cpu is rolling away yo" 😅
Even if the new microcode does fix the problem, it does nothing to fix Intel's reputation. They knew about this for over a year, did not communicate this to customers, and are now apparently rejecting RMAs of CPUs bought from Amazon and Microcenter because they're "not genuine".
People today have short memories. Especially since AMD disappointed with their recent releases. This whole incident will be forgetting two months from now.
@@westfield90 I haven't bought an AMD gpu since they were ATi because I have a decent memory. I think you're wrong.
@@ChrisL-d4c Oh they most likely KNOW about the problem, but have no idea what is causing the problem.🤣
@@westfield90 I can't wait for intel's 15th gen, AMD fanboys in shambles
@@ChrisL-d4c more like 80% of average none technical people don't know, Everyone else is all over it..
Used market for 13/14th gen is going to be filled with enough landmines to make even the Vietnam War blush...
1000 yard stare
Underrated comment
Cope
As someone who originally planned on getting a 13900k second hand to replace my 12600k in about 1-2 years seeing all of this i hope that 15 gen is fixed and on 1700 at this point 😅
@@shadowstyle3139 It's unlikely to be. 2 gens of CPUs at most...
I just updated to 0x125 yesterday and got a roughly 5% performance _increase_ (in benchmarks) and slight temperature drop on both my 13600K (Asus B760) and 13700K (Asus Z790). At full bore, both barely kiss 1.3V now.
Of important note: I undervolt both systems as much as possible before performance loss, no overclock. The new microcode and BIOS seems to respect my undervolt more strictly than before.
I'm waiting until 0x129 (or later) has a full release. Since my CPUs aren't having problems, I don't want to update them with beta microcode.
Mine i9 13900kf is barely doing 1.250v with oc to 5gzh and pulling around 195w with undervolting offset at 0.115v without any bios updates since I got it over a year ago
How do you undervolt? What settings do you have
@@raiku9161 Go in to the bios>advanced cpu settings>set the power limit to 256w>then go to core voltage set as adaptive>set offset mode to -
Then try the offset -0.050 if is stable try go lower like -0.075 or even -0.100
One comment: 13900's and 14900's have a wide range of VID settings depending on the quality of the silicon. The performance impacts are expected to be much more significant for the chips at the top end of the scale of 'how much voltage they need' than the ones that win the silicon lottery. The better cpus might not get capped at all since they are down 0.1v in a safe zone anyway. Might want to test this with a few more CPUs unless you know this one to be on the high VID side.
The Intel update specifically limits the voltage to 1.55. There have been testers who know that hwinfo CANNOT report these voltage spikes, like the hardware overclock guy, who uses oscilloscope PHYSICAL VOLTAGE MONITORS, which is the only way to see micro spike voltage. Which was going over 1.55 before patch. You cannot save degraded chips, they are permanently degraded, same with silicon lottery, and bad chips will NOT BOOST to the max especially with voltage limits.
My sympathy is with the poor devs who had to make and push out this microcode so fast. And the terror of whoever had to QA it.
It wasnt released overnight. Intel knew about it for a long time and probably knew how to fix it and how much not to push. Besides that we still dont know if this will fix it.
Bold of you to assume they QA'd it 😅
@@coolvinayExactly, no one can tell for sure, the main point of this patch was to make the CPU's last long enough, could very well fail in the span of the next 4 years; for example these generations won't retain second hand value and it is debatable if Intel would keep the suggested prices or cut them down.
It is already a worse deal based on the amount of uncertainties.
relevant take, im guessing microcode revision targetting specific variables without losing performance or functionality is not just a trip to jiffy-lube lol.
At least those guys know they aren't getting fired
I've had zero issues with my 16 month old 13900 non k. It's about 5% slower than a 13900k, but it's been rock solid and stable. I think the slightly lower power profile has kept it relatively safe. I'll probably still get the micro code eventually, but I'm not too worried, not to mention Intel has extended the warranty for my CPU until April of 2028. I'll probably be upgrading my CPU in 2 years anyway, but a nice plus.
You might have a cpu that wasn't affected by this ?
@@chillnspace777 its a non K, meaing not overclockable, meaning it doesn't have the same power issues
Non K (locked) versions of the CPU are fine
I have the same CPU. Never experienced any issues. Hoping I dont.
@@brotherrogue2310 derp, can't read
I’m rocking a 13700F for the last 7 months and I have zero issues as well.
We do not owe Intel any thank you for fixing what they broke in the first place.
Agreed
for someone who's i9-14900KS that's been working just fine, I owe them a Thank you for putting out a fix that will PREVENT said CPU from crashing. You sound bitter asf.
@@stefanokerster9844 AMD fanbois trying to drive more business to AMD, that's all...
@@DingleBerryschnappsimagine being a bootlicker for a company
@@intox69No. It's a completely justified statement.
I have an Asus z790 dark hero with an i9 14-900K. It is brand new (June 2024) and i used the 1302 bios version since day one. Everything was working extremely well. My voltage was around 1.330, and my cpu never went over 70 C.
This past friday i decided to update to the new bios (just for the new microcode).
It is not good. For the first time since i built my pc, the cpu reached 80 C. I even got an alert from my AIO. The voltage change to 1.455.
I use it mostly for gaming, so when i tried to play some Warzone, my cpu was getting extremely hot. Before the update the cpu stood around 65% to 70% TOP. But now, for some reason it was reaching 75 to even 82 C.
That's when i got this alert from the AIO telling me my cpu was too hot.
So i decided to just go back to the 1302 version, and now it is back to normal.
Sorry for the long post and my bad english, just wanted to share some info about this.
I just did my 14700k which was running 5ghz P cores and 4ghz E cores, 1.12v and 175w during CB r23 before the 129 microcode. Updated to microcode and my 1.12v jumped up to 1.35-1.42c and 215w during CB r23 with a 2k performance loss. Went back to the bios before and still running super cool and better volts and watts and regained my 2k in my CB r23 score. Either I got a golden sample or running it this way is the best solution to those at intel that think they need to push clocks so high to compete with AMD that they forget about longevity of the average to hardcore PC user. Before this I was using an 8700k so keeping the P cores at 5ghz and E cores at 4ghz is a massive step up for me. These CPUS never should've been marketed at 5.7ghz to 6ghz on the performance side as intel chip isn't as good as TSMC who makes AMD's. I'll stick to my setup and not worry about the issues for running it too fast, too hot and getting degraded over time.
Similar situation. For my case, going back to Old Bios does not help. My machine has now been destabilized. I'm not sure what causing it T_T.
@sugardaddy9721 even tho u went back to a different bios version ur microcode stayed the same
@@def-po8tu i thought going backwards on Bios revert the microcode
@sugardaddy9721 no, it says it in the warning when flashing bios that the microcode will update and cannot be reverted
Small side note here. In physics Power is always produced by multiplying voltage times current or amps (ampere as the french say). So, we have V * A = W(atts).
That means, there higher the voltage for a current is, the more watts you have to cool.
Before I RMA'd my 13700K, I was watching Buildziod's voltage analyses video and I thought to myself, what if I fiddle with the resistance and voltage directly in my VR-settings? So I did.
I set a upper voltage limit to the VR (voltage regulator, which you can fiddle with since Z690) to 1.380V, which means, my 13700K got an upper voltage limit to 1.380V to run at stock speed.
I also lowered the resistance settings in the VR-settings from 40 / 90 AC to DC loadline to 10 / 10 AC to DC loadline. And I set the current limit to Intel's specification to 307 Amps.
With benchmate I got my 13700K from 29600-ish point in Cinebench23 multicore point to 30200-ish points at stock and 30962 point with a +300MHz overclock and +200MHz on the P-cores and E-cores, but I had to adjust the voltage limit up to 1.450V.
However, within Intel's spcification and my self-emposed voltage limit, Cinebench23 did'nt even get remotely close to the power limit of 253 Watts. Highest measured was 233 Watts and the highest core temp was at 82°C. That is a world away from the previous 104°C.
I'm running my 13700K paired with a Z790 MSI Gaming Plus WiFi motherboard and a DeepCool AK620 air cooler, at stock ratios, 110/110 AC DC load lines, ICCMax at 307A, IA CEP=enabled, and a global negative voltage offset of -150mV. This applies to both the CPU core and Cache, lowering their VIDs. PL2 is capped at 188W, max temp in CB R23 reaches 85C with 23C ambient, R23 score is 30200 pts, the highest I've seen is 30313. VCore during this multi-core 188W R23 load is around 1.18V. If I set PL2 to 253W, the max draw is 215W and the score is around 31K. Microcode is 0x123, currently running the latest stable version BIOS for my MB.
Max VCore spike I've ever registered (HWInfo is running 24/7) is 1.338V, during light low-core usage.
People online tend to mostly undervolt through lowering the AC load line, but my personal experience is that I get the best results with keeping AC=DC=110 (those are MSI's "Intel Default" settings), and then applying a negative offset. No instablity at all, I've successfully passed all kinds of tests such as y-cruncher, OCCT, Prime95, Cinebench R23 and 2024, TimeSpy Extreme, Intel's Diagnostic Tool, and of course everything is stable during regular usage and gaming. I can't really be more happy with my CPU and my configuration, hope it stays that way in the long run.
@@LooneyTo0n Common sense and deductive reasoning in one post. Almost like you have done this before. =)~ Cheers!
@@tysoncoffman7562 I try my best! I've also put in many hours spent researching the topic and experimenting with settings.
Purchased the 14900k in Feb 2024 from Microcenter. Was having non-stop crashing during that time trying to play Helldivers. I saw your video about the power settings and that solved the issues of crashing, but still had program stalling. Updated to the 0x125 yesterday on my MSI board. Still having issues with program hangs during high loads. Tried running Cinebench and explorer will hang and crash. It's insane. I will be heading to Microcenter to replace the processor on the weekend. I hope that fixes the issue.
If you have still having program stalls, then you're still getting computational errors from the cpu, which means your CPU is really sick and still dying even with the new power settings. You should definitely RMA it when you can.
Are you using a high power profile in bios? My default power profile is basically unlimited power on MSI board it is unstable, set to intel defaults and helldivers etc is all good.
That fix might as well only affect future damage
@@damianabregba7476 Not really. It just extends the longevity and delays the eventual failure for a time. It's still defective and rust never sleeps.
If you are already seeing this, the CPU is toast, only way is to RMA
13:30 “there’s 28 threads in a 14900K” small correction, there’s 32 threads in a 14900K.
Correct. I was still thinking 14700k since that's what we were just testing before the 14900k
4 threads were fried in the process of testing, turns out the degradation is going at slightly faster rate than originally anticipated. Still good for 7 more tests, remember NO REFUNDS
No, it's probably 28 threads at this point. lol
@@rustler08The cores that reach + 1.5 volts are the P-cores and the ring voltage is the same as the core voltage.
The ring can't handle the voltage, the cores can, i learned this from leaks from Intel engineers in Moore's Law Is Dead, they said that they know since Alder Lake that the ring can't handle that much voltage.
Considering that Intel wanted to launch Raptor Lake fast, that explains the quality control issues.
@@roh_son 24 physical cores and 32 virtual cores (threads).
I hope someone comes up with a stress test that that will difinitively show if a cpu is bad or not.
Just launch an Unreal Engine game while cpu is OC if "memory error out of video memory" and the game closes. It's still a cpu fault. If it BSOD its a cpu fault.
Can crash-reboot be this too ?
Someone posted that if you try to instal nvidia driver 5-10 times and got an error, then the cpu is bad
just run it stock watts for the sku you have. if it doesn't deliver the advertised turbo performance it's trash. technically or marketing error. upto no good clause.
@@glitchered That doesn't demonstrate whether any 13th/14th gen CPU has been damaged or is suffering from this issue, all that will do is demonstrate that you lost the silicon lottery. A better test is repeatedly performing any decompression task larger than 300MB 5-10 times. If it errors out, or the time remaining goes into hours or days, you have an affected CPU.
I bought my first Mac. Macbook air M3 Base, only 8GB ram and i am choosing to complete most tasks on it. I cant believe how much better everything is on this thing.
I don't know guys how you made the shot at 16:04 but it looks just great. With the ring sliding out of the view and slow camera zoom. Perfect.
I would honestly take a 5% performance hit if that means there really is a significantly lower chance my CPU will burn itself down. But yea, even if intel announce a long-term fix, we wouldn't really know whether it works until maybe months or years later.
The CPU’s are dying from Intels dumb stock bios setting which pushes 1.5 volts to two of its cores. If you change its stick settings it will never degrade.
On stock settings my cpu was pushing 1.495 v on the vcore
With the 0x125 microcode update and a slight 300 mhz underclock paired with better ac/dc loadline settings it never gets beyond 1.14v and is very stable
They dropped the ball with their defaults at launch
This is what gets me. I've got a 13600k and need to get this upgraded at some point, however I don't know if I want to stick with Intel in the end. :/
they should offer EVERYONE a new chip to replace it with the old one. & then u can sent it back after you recieved a replacement. 1st Faulty product or working product that was ran out of spec by intel. 2nd false marketing. 3rd its already semi broken for some people since the chip degraded & it will not guarantee a long life expectation. a new chip will may last 6-10 years while the old chip can break down in the next 1-3 years.
i wont accept 5% decrease. false advertising is a big deal. getting what you pay for matters. apathy like yours leads to capitalism being corrupt.
I just went back to my 12700KF from a 14700KF. Amazing how much more reliable and cooler the 12th gen chip is.
I just love the fact that 12th gen is considered "cool".
And that's saying something. 12th gen was already known for being thermonuclear
agreed. My i9 13900K hits 100C instantly on Cinebench. It's insane.
The 12700K is objectively speaking the best i7 from the last decade in perfomance per watt, it makes the Ryzen 7 7700X look very dumb.
@@Grievous- Cool is a relative term.
A 95 degree day is cooler than a 100 degree day.
See also: "Cheap mansion", "cheap supercar", etc.
Let me know if you're still (somehow???) confused.
aaaaaaand since this "fix" is going to require months of failure data to see whether it's had any impact on reducing the rate of degradation - ONLY FROM THIS POINT IN TIME ONWARDS - this issue is guaranteed to continue being unresolved and a total PR fustercluck for Intel for months to come
This.
Never been so happy to have ditched intel. Degradation won't continue at an accelerated rate for the affected CPUs, but shortened lifespan is a guarantee.
quote: "...because you are not using a genuine one, you are using fake cpu...." as if it was so simple as manufacturing a pair of shoes...Someone went too far with its statements. Too far.
I undervolt my CPU as many people watching this video needed to do since the heat the 13th and 14th gen CPUs can generate otherwise.
My undervolt settings do not seem to work as well as they did before. I was happily undervolting by an offset of -0.070v prior to the new microcode. It was just a simple and modest undervolt by setting the negative offset in the ASUS bios and could run the Cinebench 30minute stability benchmark just fine. It worked for well over a year.
Now I can only apply a -0.050 offset. It does seem to still keep the temps down but I'd love some additional footroom just in case. Cinebench doesnt go above 95c in the 30 minute test which is similar to before. Maybe its just that the update addressed some of the exxcessive voltage and I no longer need as high of an undervolt?
The next microcode in line would actually be 0x12A, then 12B, etc, they're in hex notation.
So 0x131 would be 13 versions away, and not just two.
Isnt 131h - 12Ah equals 7?
If we keep the progression, it would be 0x12d, as it was a difference of 4 between 0x125 and 0x129.
And 0x131 would be 8 versions away, not 13... ?
You must be fun at parties
@@AutomotiveConcepts Hey, this is a nerd gathering, get your toned muscular a$s outta here!
@@Winnetou17 You're right, it's 8 and not 13. Should've gone to bed instead of trying to calc in hex that late 😄
Very interesting Data. Just goes to show, the high voltages weren't at peak power loads, and not peak workloads (ie. multicore). And with the minimal performance drop with the new patch, you can really see that these high voltage spikes weren't really necessary. They didnt' give you more cinebench performance, they gave you "better performance while loading" ie, when no one really cares or notices. These voltages were to hit clock speeds that were largely marketing bulletpoints, which minimal performance benefits. However what they did do is degrade the silicon. Hopefully Intel learns their lesson.
Yeah they learned they need to slow the planned degredation of the next generation by just a few percent so the chips don't fail until after the end of warranty.
With a very old BIOS my 13900KS never exceeds 1.4 Volt. So, I'm definitely not affected. The CPU runs heavy multi threaded and also heavy single threaded workloads at a daily basis with no issues for 17 months now.
@JayzTwoCents MSI just dropped another updated BIOS today (8/13/2024). You may want to retest with the newer 0x1 129 BIOS. I retested my 13700K on an MSI Z790 TOMAHAWK WiFi6 DDR5, and I saw a significant boost in the Cinebench R23 multicore score. MY BIOS settings are set to Intel Default Safe, XMP1 bringing DDR5 to 6000mhz, and CPU capped at 253w.
Even i a never had an issue with i7 14700kf on msi z790 tomahawk max wifi, should i update my bios ?
I remember seeing Falcon Northwest ads in 3DMark 2003, lol.
Didn't it say something like, "If you were running this on a Falcon Northwest PC, you'd be done by now"?
Damn I’m glad I went 12th gen, sucks I can’t upgrade but it’s ok, I’ll probably be ok until the next fully build in 5ish years
Why can’t you upgrade bro? 14900k is a beast hopefully get some discounts given the brand reputation hit.
@@GiljrgI'm sorry, are we even watching the same video?
@@Giljrg kek, this guy.
😮huuuh? running z690 with 12600k, 13600k and 14600k
or did i change motherboards ?
This isn't like a GPU where you could go to your closet, pull out an R9-390, and have a usable (but hot and slow) experience while you wait for the RMA+shipping both ways. A person who has the means will go out and get an AMD MB with a 9700 or 7800x3d and do a system transplant. *They might not go back to intel ever.*
The 'old' intel MB and when (if) they get the CPU will end up as someone's "gift"
As someone regularly taking care of other peoples systems the thought of a wave of used Intel cpus is pretty unnerving.
considering amd is running into issues on their new chips this is nothing new to any company, both amd and intel, nvidia and so on have all ran into multiple issues
@@yulfine1688 i am pretty sure many people choose Intel because of their reliability, plug-n-play superiority compared to the mess from Ryzen memory bs, especially their 1st gen.
guess now Intel is expensive, hot-power hungry, and unreliable. surely a winning combo for any product LOL
@@MuhammadRidwan-pe7ny intel has always been on the power hungry side, heat has been a mixed bag, for the rest that hasn't been the case for almost a decade now, amd has done a generally good job.
Also they have been mostly neck and neck with intel at times having a slight edge until the x3d chips which sometimes have an edge.
People stick with intel because they're large and well known its a comfort pick.
The average person is also usually stuck with intel because most laptops and so on are intel, though we are seeing more amd laptops.
meh. Y'all acting like AMD has never had its fair share of f ups. Funny how all of those are magically forgotten about right now. All Intel has to do is release a good chip and all of this will be forgotten except by terminally online people who tie their identity to a corporation. Everyone is jerkin AMD right now, but they really did used to be an awful brand. They made good chips and now everyone likes them. The idea that Intel can't recover their reputation is just ridiculous
Alright 16:03 how long did it take you to get that shot lol. If that was cut from two pieces of footage that's one of the smoothest cuts I've ever seen H O L Y
Intel today is not the company that pulled their halo pentium chips for the famous alu bug. The company had better leaders then who wanted to please their customers not the stockholders. No corporation deserves our loyalty, plain and simple. Corporations today are money making machines for the elite stock holder not you the customer. Appearances and pr is more important today than a lasting product. Blame it on ideas like manufactured obsolencence and closed systems with no right to repair. I would apply the same for AMD also. Brand loyalty is not a good long term strategy for the consumer.
I’m a stock holder. I don’t have enough share for the dividend to be shit but I would still prefer good products. I invested in Intel because I use their products. Bought some more of this cheap stock too. I don’t recommend that
Imagine if it was actually easy for normal everyday people to buy stock.
Market would look nothing like it does today.
@@thepopeofkeke Obviously the typical individual stockholder has not much voting rights to influence. I was talking about the institutional investor, the Blackrocks and Vanguards of the world and the elite investors they represent.
@@AceStrife Stocks are a democratisation of company ownership. But it has become too complicated today.
@@AceStrifeThe stock market shouldn't exist imo. It's one of the many things poisoning so many good things with greed.
MSI of all companies updating their bios first? That's a surprise.
The real surprise would be if Gigabyte did it first.
@@danielgarciasalorio8646 I cant speak to the new platforms but for am4 gigabyte is actually on top of the game for motherboard updates. My first gen am4 board still gets regular updates and normally before anyone else. Its possible that they are very different teams though haha.
ROG STRIX B760-F GAMING WIFI BIOS 1662
Versión 1662 Versión BETA
12.84 MB 2024/08/08
"The new BIOS includes Intel microcode 0x129
MSI has been trying to step up their game lately. I wish motherboard manufacturers would start competing again, stagnation is real, basic features are being sold at a premium these days and it sucks
When freakin nvidia comes out and says “don’t look at our gpus it’s on intel” and they’re completely in the good there, it’s a fucked situation.
When Jay said "I love Computers" i immediately imagined him in "i love refrigerators" parody ad skit
Im on old microcode and no plans for updating bios ... Tuned performance and thermals myself ... ACDC Loadline 28/55 LLC Medium , core offset -0.055 V , 253W PL1/PL2 ,CB23 score muti 39840k on gibabyte Z790 gaming X AX @ 13900K , no thermal throttle , stable as hell passing Ycrucher ,and CB 10 minutes and 30 minutes runs ...VID MAX never exceed 1.360V , VCORE MAX pek 1.380 .Amamzing performance and low temperatures !
Yo, Jay you are 1 of the few Pc UA-camrs I actually trust. I trust your judgment I know sometimes people can be wrong. But when it comes to you man, I know that you are here for us, you try to do what’s right as far as information goes. Thank you for helping us!
@@DingleBerryschnapps I bet you trust Linus Tech Tips and are a fan boy lol
Jay is fun to watch for fun projects but for straight up fast solutions, I would watch FRAME CHASERS. FC's found and fixed this whole Intel issue last year. It's just too bad he didn't have a massive following when he had the Intel fix.
Remember, we just want strong Intel and AMD, and if it were possible, others, competition is key. Yeah so I think it’s reasonable that people expected a drop in performance, without being sure it would.
!@#$ Intel. If you still trust Intel OR Microsoft after these current debacles then you DESERVE to get shafted.
@@conza1989 The 13900K multicore is faster than a 32 big core Threadripper 2990WX, i don't know why people would complain about a 7% drop in perfomance with a 100 watt reduction in power consumption...
That is still, way, WAY too fast for a consumer CPU, unlike Threadripper, extremely fast for games as well.
I advise people to limit the power in bios and wait till more testing is done, with these rushed stuff you never know they will not pass throught the usual QA and will end up fucking up.
It's weird, because actually, for my 13900K it was more stable without limits and overclocked than with a limit, at least while running Cinebench R32. I tried the limits when people started saying to put the limits, and it was bad on Cinebench and had crashed after a few runs. Never had a problem with games or apps just while torturing the CPU on Cinebench. Without limits runs great like it has since I bought it over a year ago. But after I tried the microcode 0x125 a few weeks ago and its stable no matter what I try! I get a score of 40200 in Cinebench R32 without limits and I can run it for over an hour without issues! so, I am not worried. and probably won't even try the new microcode since it's so stable now since the microcode two weeks ago. Anyway, planning to change it to a 14900KS soon.
@@rafaelrosado5454 CEP cripples performance... but the others are needed.
Just limiting the power will most likely do jack S#!T, since the main problem is the high voltage. Which is present when a single core boost happens. Which, guess what, is when the processor runs at 30-50W. When at 200-300W, the voltage is actually lower, just like the frequencies are lower at all-core workloads vs single core workloads which can boost to 6.0 GHz.
@@Winnetou17 if it is not broke don't touch it, ask any it guy.
so if your cpu is experiencing 0 instability just chill and don't fall for the fear.
What is your settings? Can you please tell@@rafaelrosado5454
A lot of the lower quality high end Raptor Lake CPUs can request more than 100mv extra, compared to a good sample, and those will be the ones that see the most degradation. Better samples will have lower VID requests, so the even the worst spikes on the old microcode might not even damage a good sample. The question is what voltage bins had VIDs high enough to be vulnerable? The worst 10%, 20%, 50%?
thanks for the update always look forward to a video from you guys
Seems mostly margin of error stuff. The real question which we'll take time to get the answer for is whether this will actually fix the issues or not.
Jay also only really tested the 14900k. Performance deltas could be different on the other stuff.
They trying to fix it for last two years... so this new "fix" released in the same time when AMD showed their new stuff looks to be PR fix not hardware fix...
I have the MSI Z790-A Max Wifi and I have no new updates to Bios, only a beta from 5-30-2024.
Same
It’s only for specific boards from msi
Same ,I thought maybe I overlook but show that old beta one still ,maybe there was a issue with it . Who Knows...
Yes. To give them the benefit of the doubt, I sent a support request to MSI for a beta bios (in their forums they were saying you have to ask for it) but they answered, "sorry, no bios for the motherboard we sold you." This represents my first and last customer service experience with MSI. Live and learn.
@kpm4620 my msi board has a new bios updated posted today but still says beta. Not reassuring.
I have a 14900k on Asus z690-E and I've had it blue screen, game crash and chrome crashes in the last few months. I will wait for x129 update to come out officially and see if it makes a difference for stability, otherwise if my CPU is degraded I will go through the process of RMA so I cant wait for that headache.
GL, If it has degraded that would be the best. Not the same issue but my 12700KF & 12900KFs got stable e cores off, turbo boost off (causes some games to crash), switched to Linux (been gaming on Nobara past few months) Asus Z690 TG Wifi. Same machine(s) would BSOD previously just the 12700KF months ago in Firefox not doing much it was the last straw for me to try more things. If degradation has occurred would RMA though.
How long have you had your cpu for?
@@Crynomical 12/5/2023 date of purchase.
@@HOMERUN1379 I’ve had my 13900k since December 29th/2022 and no problems here yet
@@Crynomicalsame here zero issues
If running a Windows Preview version, it might be a good idea to unable encryption/de-crypt drives before updating UEFI.
Bitlocker might send you on a recovery prosess for unlocking your drives.
Might be relevant as bitlocker "light" gets enabled on Home versions.
Worst for the voltage control are quickly switching workloads. All-core load are not most stressing for modern CPUs (i think since Zen3/Comet Lake), because they just won't push clocks as high as in single-core loads, so they don't need as high voltage, even if they consume most power they can.
There are tools like CoreCycler to test switching workload behaviors.
Just remember, if not for Steve from GamersNexus Intel would probably NEVER say anything about it, nor made an update
Was he the first to notice this?
@@GreyDeathVaccine Can't be 100% sure, but it's his investigation which has unraveled it to the big public
The Tech Jesus😅
@@martinxyz Not even close to the first one to notice it. Every medium-large game dev knew because of the sheer volume of crash reports. Ive been having issues with my 13900K since the start, and I ran into specific pages for 13th gen CPU related crashes on every game I tried to resolve them in. Quickly became clear it was my CPU that was the issue.
We've been talking about it for well over a year, but happy to see it finally get into the mainstream.
@JayzTwoCents 8:28 Please move the lower end of such Graphs up a bit in the future. If i pause the Video, the labels are just under the youtube video controls that are visible all the time when paused.
I work in IT - No offense but given Intel's frankly shameful response on this issue I won't be "canary" for it's microcode update. Given it's also flagged by MSI as "beta" I'll wait until they've *fully* tested it.
It seems fine voltage spikes are gone, which was the main culprit.
Im running a 13600k. I recently updated to the new microcode and i started getting higher temps in benchmarks. I have seen a couple people report that MSI has boosted the lite load in the bios to still draw more power. I verified this myself and saw lite load was at 18 when 12 was the old default. Turning this down can greatly improve temps without hurting performance.
It would be super helpful, if you'd be willing, to put out a video for those like myself who are now having to learn how to update bios and adjust settings in bios because of the Intel blunder, showing us how to do so safely. I know with ASUS boards you can use AI suite 3 but some people say I shouldn't. I'm having a hard time learning properly with so many contradicting opinions, it'd be nice to see some of the bigger tech youtubers like yourself put out videos to help those of us who are much less read on this subject.
I haven't been able to download the MSI update all day long 😢
If you have an MSI, don’t bother with the update. The update essentially solves nothing. Permanent processor damage has already been done.
Contact MSI about complete system replacement
@@ronaldckrausejr7762just like all the updates they said if its damaged its too late
What motherboard do you have? its not release for all of them yet.
It’s only for specific msi boards
I am waiting for the dust to settle before I download anything. No more beta testing for me. I have other computers until this is resolved. Luckily all I have done is install the os in the msi computer. So hopefully it hasn't been damaged. The other one is a gigabyte and I haven't bought the processor yet. May wait a year at the rate they are going.
What I can tell for sure is that Jay is not a fanboy of ASUS mother boards and probably their other products 😂
Asus were the WORST offenders with allowing ultimate powa to the CPU.
Got to give props to Intel on the RMA process, no questions asked after they read instability issues. They also overnighted the product to me once they received it.
Great job, Jay. Thanks for banging this one out real quick.
Did the latest update on a friends 0x1 129 for I9- 14900K CPU using an nvidia 3060 video card and couldn't even get into the bios after the update. He didnt want to try NVIDIA GPU Firmware Update Tool for DisplayID so Installed the bios update from May, prior to all the intel updates and now it runs fine.
Is it supposed to be difficult to figure out how to update the microcode, or should I wait until intel details how to do it?
Asus states in the BIOS descrption "The new BIOS includes Intel microcode 0x129 and adjusts the factory default settings for the non-K processors, enhancing the stability of Intel Core 13th and 14th gen desktop processors". So, this would be useless for anyone with a K or up processor?
Good question, wondering the same
Hi, So from my testing, my CPU was always crashing randomly with the default settings ( I have 14900k ). He said that if your CPU was crashing then its already dead. I don't think this is true at least at first glance. Now I do not know if the CPU will keep degrading, but I can tell you that after the 0x129 microcode update of my BIOS which was 7 days ago, my CPU has not crashed once while gaming using the default settings. I have not set it below 1400V ( which I had to do before to be able to achieve stability ). So based on my experience I think that the microcode works just fine for K processors ( Although I have a Gigabyte motherboard ). What I noticed as a minus for me at least is that with the default settings the CPU is a bit hotter while gaming, but at the cost of some performance if I reduce the voltage the temperature will drop as well. Will not touch anything for now though, want to keep gaming like this and verify if I will have any crashes. Of course now we just have to wait, I guess time will tell if it is really effective, but it definitely looks better for me at least for now.
"The latest microcode update (0x129) will limit voltage requests above 1.55V as a preventative mitigation for processors not experiencing instability symptoms. This latest microcode update will primarily improve operating conditions for K/KF/KS processors."
From Intel page
My 13600KF CPU is damaged, and I just sent it in for RMA yesterday. Every Cinebench test resulted in overheating, reaching 100°C, even tested a few different cooler and the latest BIOS from ASUS, default intel setting. Blue screen randomly in a clean window. So much for the legendary “Intel stability,” hahaha
How is that damaged? You just have shitty bios settings if you're hitting 100C or a shitty cooler with poor contact for example.
that s poor on your side bro; either the cooling; or you still changed something in bios settings; i have 80 degrees on my 13700kf; be sure that ur bios is on default; and you can set anytime max cpu temp in bios; unless u bought a shitty motherboard and now wonder why ur cpu is getting cooked
@@KoItai1 i have two same CPU. Done swapping test and confirm.
Don't use default Intel settings with 13600kf, those just result in over voltage and overheating. Same with my 13600kf, use the motherboard defaults and undervolt it if you can, way better
Fixed my MSI / 13900k setup. I had to press and hold the power button down a few times after the bios update. I also applied the FW update under the bios. Once I was back up I had to reset my Microsoft Login info. All good now! Don't forget to also go back into Bios and turn on XMP.
Not to be intels fan boy, but i hear alot of "bad stories" by youtubers and read on different forums about INTEL's RMA process. I just got a brand new CPU sent from Intel, i wrote their support saying that i was worriesd about degredation, they told me to send it back and they would send a new one to me. The email's back and fourth took 2 days, Intel paid for a delivery service to come and collect the old cpu at my address, they paid shipping, and the cpu had to shipped to a different country. The cpu left my house monday a 11, and a brand new one arrived thursday around the same. Just saying, if u got any worries about your cpu, dont hesitate to write to their support, this is the most painless dealing ive ever had with any company about any returns, super nice and polite support team, and super quick as a i was expecting to be with a cpu for atleast 14 days. Side note, thx for the video, informative as always. Hoping u gonna do some OC testing on an asus mb sometime in the future with this new microcode.
Seems like it goes both ways, I read a comment from someone who is having a terrible time getting his CPU RMA'd, they wanted pictures of the CPU, attempts to reset the bios, and some other junk I don't remember and they still haven't approved an RMA for a 14900k. He said it felt like they were purposely delaying him.
Does your country have strong consumer protection laws?
Wow I am gonna try that
@@volvo09 I live in EU, Denmark to be exact, i couldnt say if we have any perticulary strict laws in place, compared to other countries to be honest.
Also i dont know if they apply here aswell, as im thinking it just falls under to the standard 2 year warranty that comes with the product. B
ut again, im no expert here, so couldnt say if where i live is a factor.
Sad for that other guy though, truely the very opposite of what i encountered.
I’m just sitting here watching all these CPU issues and RAM compatibility and I’m very glad I’m running AM4
AM4 😂
feels like intel are paying the price for throwing insane wattage at their CPUs to keep up. Crazy the difference in power requirements.
People absolutely HAMMERED AMD for doing that around the times of Bulldozer. I do find it quite ironic the way that the reverse is happening currently, and the people who were vocal then, are quiet now. I've never advocated fanboy behaviour, but people should at the very least be consistent.
LOL, no. I was one of the people dunking on AMD back then (bulldozer and especially PILEDRIVER) for 'HURRDURR MORE CORES', and now I'm dunking on intel for selling defective-ass CPUs with horrendous power inefficiency and trying to brute-force the problem. There's plenty of consistency, you're just not seeing it.
@@NapalmFlame You're a minority, I think. For some reason, in a similar manner to the brand loyalty people provide in the GPU market, it seems that people feel obliged to "Pick a Team" I can't say I understand it, but it is what it is and I do think that there is a large majority that do so.
@@NapalmFlame Bulldozer or PILEDRIVER CPU were nowhere near this wattage numbers... ZERO CONSISTENCY!
@@richardsmith9615 85% of every content creators comment section are Ai.
For Asus it seems the 0x129 is in Beta and says for non-K 13th and 14th processors as of today. (Referencing Asus ROG Maximus Z790 Hero).
Does that include voiding your warranty
I have same mobo and just updated to test it and what i noticed that there isnt anything new i tweaked most of those things manually and actually i got 35k cine score with new bios stock power and 82C on cinebench before i had all cores lcoked 5600 and 4400 with manually tweaked voltages and i got 71C and 39700 cinebench score lol
@@gvido2923 Good to know. Thank you for sharing your findings.
Find it strange that msi bios should have an help on -k models and not asus, strange? isn't it same microcode?
MSI MPG B760I Edge Wifi - updated my bios due to the Intel Microcoding update. After flash system booted fine. All CPU settings and ram settings correct. Later in the day after work loaded up some games and no sound. Reinstalled all audio drivers and still nothing. flashed back the bios to 7D40vA8 and still no audio detected by Microsoft. Even reinstalled all Motherboard drivers, windows 11 up to date with supports. Created a ticket with MSI but I fear Audio Chip is FRIED! Keep up the good work Jay, you did warn us to hold off on updating lol
Oh wow 😊
Just subscribed, such a straight to the point video just about the facts. Its very much appreciated so thank you. :)
Ooof that cyberpunk drop, paid for 14900k and got 14700k.
o wow 6 fps you'll never notice
@@djcetra will you notice if its money? A 14900k is like $550 and a 14700k is around $380, so you lose $170 worth of performance in one update, that sucks. 🤣
@@djcetra lol 6ps is life or death for some people ya know 😂
Won't the 14700k also see a performance drop after the microcode update though?
@@alison9870 likely, Im sure every single tech tuber will eventually test them so we will know post haste. 🤣
I might have missed it, but was this run on jays golden sample?
Was anybody else glued to 9700x and 9700x PBO results? My eyes just kept snapping back to that with each new graph.
Yeah AMD 9700X is way better overclocked than stock. See Der Bauer video. 5400mhz all core and it will pass Intel. So IPC is a tad better.
It did CB23 -> 24000 points, my 13700K with E cores disabled did around 23000 points.
Actually, I was watching the videos yesterday about how AMD's been lying about the performance of ALL their new CPUs and thinking, they suck too.
Thanks @JayzTwoCents for covering this so quickly.
Jay finally delivering the content that community needs and is actually the only creator that i found that seems to do it right and early. Big props!
If there will be further microcode updates for this issue, see if you can find even more scenarios on how to test/replicate the 1.6v bursts and new behavior.
not all of the MSI boards have the BIOS update yet. As someone who has a Z790 Edge Ti Max, that update is not yet available
Same with my board PRO Z690-A WIFI.
"We have sold you a car with the front wheel axle sawed in half. You can lose steering and crash. We have found a brilliant solution: a firmware update that limits the speed if there are holes in the road or if the shock absorbers shake too much, also the speed in general.
You can still end up going down a cliff and die.... in that case we'll see what to do on a case by case basis if we feel like it. Don't complain or we won't help you anymore."
This is great reliability and customer service. Who knows why they lost 40% of their stock value in 1 month and 50% in 6 months?
I run my 10900K exactly the way Intel is saying is not safe, no power limits, no c states. Still rock solid, same OC as day 1.
Intel has totally ruined their reputation over this scandal. I've been an Intel Andy because of their reliability, compatibility (no weird XMP issues or CCX concerns), and stability. Now I am not sure exactly what my next CPU will be.
I know everyone's pissed at Intel, but if they can fix this it is excellent news, and we should all be rooting for them to rebound. If Intel fails and AMD has no competition, CPUs will become extremely expensive, and progress will slow significantly.
i mean; 5600x was 30% faster than 3600x; 7600x was 35% faster than 5600x; 9600x was... 1% faster than 7600x ? Thing is intel was always good on platform jumps; which should be the case for arrow lake too; then they were always like filler gens or refreshes after the new platform; maybe this will change
can't fix already degraded CPU's and lots of ppl in limbo with useless CPU's
@@LiLBitsDK ye; can t fix degraded cpus; that s why you have warranty for 5 years in the case ur cpu gets degraded; lol
Already updated on the Gigabyte AORUS board for myself. Glad to know I am on the safe side at least for now. Have not had any crashed or instability on my 14900K.
Thanks for the video. Would have liked to see a frequency comparison for the Single-Core stress test workload, as I believe you only provided the 10 minute all-core workload chart. As you stated, the single threaded is where you would notice differences in frequency the most (and see if 1-2 cores can still even hit 6ghz anymore), and would have a significant impact on gpu-limited games.
I'm sure my half fried defective cpu will be fine now.
Just sitting here sipping tea with my 11700k with zero problems and no “efficiency” cores
...and yet Intel 11th gen chips have one of the highest RMA counts prior to 13th/14th gen. As is often the case with hardware issues, "you got lucky".
as an extra benefit you can warm your tea by putting it next to your 11700K cpu
@@EightPawsProductionsHD No, because people found that with the 11900K they got what this guy has.... An 11700K, and got pissed off...
Dare we say when AMD was shitting the bed with $100 2600x and $125 2700x, AMD boys was telling people on reddit to buy 10400's and B560 motherboards. Then the same people saying that crap was on Discord building their own 10400 B560 rig LIVEstream and finding out their top M.2 wasn't being used... After telling 1000 or more people to buy that same setup....... It says so on the b560 specs and in the manual.....
*_"Just sitting here sipping tea with my 11700k with zero problems and no “efficiency” cores"_*
Just sitting here smiling with my 13700k with zero problems, because I'm smart IT guy, and had it slightly underclocked from the start, instead of allowing the motherboard manufacturer trying to blow it up with its settings in an attempt to score great in reviews.
It scores 46,494 in multi cores benchmark vs 23,963 for your 11700k.
It scores 4,362 in single core vs 3,406 for the 11700k.
Google: "Puget says its Intel CPU failure rate is lower than AMD Ryzen failures"
*Chart*
AMD Ryzen 5000: around 4% failure rate
AMD Ryzen 7000: around 4% failure rate
Intel Core 10th gen: around 0.5% failure rate
Intel Core 11th gen: above 7% failure rate
Me chilling over here with my 18 core 7th gen i9 pushing 4.5ghz all core with no dumb efficiency cores 😎
whats the reason for 13700k outperforming 14900k in cyberpunk?
Fewer cores/threads maybe. Could give more thermal/power headroom in workloads where you don't need as many.
im more impress that the 13700k beat the 14700k 😅
Test should also include the version with uncapped mobo settings, because that was the default running mode for most people pre shit hitting the fan. There were a lot more noticeable multi-core performance drops when switching form uncapped to Intel Extreme Profile, which lowered power consumption about 60-70W average (sometimes even more).
It looks like the saga is coming to an end. Now i would like to know a post mortem. was the issue blown out of proportion?
The amount of power the world is wasting by running Intel atm is crazy. I get that people don't care in areas where energy prices are low but damn, those suckers produce some heat.
My 14900k using 13w at desktop 70w in games, my 4090 on the other hand is what heats up my house 😂 cpu power draw really a drop in the ocean compared to GPU
The proverbial neighbor complaining about another neighbor's sprinkler system, while filling up their own pool.
What about I didn’t know about this problem till after I bought it 😂
@@Giljrg LOL bruh my 4090 bakes tf out my room I hate it lol (not the gpu the heat)
The "burning" of the intel chips will continue until morale improves!
Like this comment if your 13/14th gen survived until now.
i have 4th gen i7😀😀
I got a 13600kf that I’ve had undervolted since I got it
Mine was basicly dead in a week
Idk mine now has small hiccups after playing for a bout 20 minutes 😢
I have a 13400F that hasn't had issues so far. It's also true that we're talking about a rebranded and locked last gen 12600KF, so it shouldn't be affected in the first place
Aye right falcon northwest, bet ye were big in the gaming scene building 486dx2's
Yep - 386SX & DX as well. Year 32 here.
@@Falcon_Northwest Wild boys.
I just did what Buildzoid did and limited my CPU Vcore input max to 1.4 volts (plus a couple other LLC tweaks) and everything's been stable since. Granted, it doesn't boost over 5.7GHz anymore, but I'm still getting mostly the same Cinebench results as I was with the unlimited/4096A power limit settings, so I'm fine with not getting 6GHz boost if the overall performance is the same. I'll still check out the 129 microcode though when ASUS releases it for my board
did the max 1.4 volt thingy to. Just to be safe i put the pl1/pl2 at 125/253 watt and set max amps to 307 (i have a 13700k)
@@Ruud_Gyou also use 55 for the AC/DC?
@@haies09 i didnt touch those settings
@@haies09 I tried that on my Asus ROG Strix Z790-A board but it completely killed performance. Like I was only getting 10% of what I should've been getting. I think that setting is specific for Gigabyte boards
@@Ruud_G I used the Intel extreme profile before changing the Vcore limit, so my PL1/PL2 is set to 253W and max amps is 307A
CyberPunk 2077 is what gamers call "a resource hog" Fallout 4 was the same way, constantly tracking NPCs
Is there a tool to determine that there is or is not damage? The Intel Processor Diagnostic Tool does not clearly show that it is or is not checking this.
How do you want to check for a failure that so far has had about 4-5 different reasons. Run some performance intensive workloads or run some benchmarks and see whether it crashes or not.
As far as I know, If you run a benchmark and look to the frequencies and voltages of each core (you can use hwinfo or cpuZ). If all of them are in the expected range for performance or efficiency cores, them there is no damage.
However If one is getting hotter than others or If your system crashes onde in a while, them you have a problem
nope, unless u experience crashes or bluescreens with stock settings. it may already degraded - no chip is damaged - they all degraded to a point it gives out errors & then ppl call it "damaged chips"
I have a bit of professional experience in semiconductor manufacturing and reliability/damage. The physical damage is going to present as erosion (causing leakage) and / or shorting between transistors and traces. There are systematic electric testing that is done in QA prior to packaging that can find such problems, but I don't know if these tests can be replicated purely in software post-packaging. Since damage is not binary, the quantum well errors must be statistically calculated over a fairly large number of test runs. With 4.2B transistors, this would be a long and involved test, but again, it's unclear to me if exercising specific cores and transitions is possible without physical access to the test pads on the die.
If any semiconductor test and QA engineers are reading this, I'd love to get your opinion.
Run Prime95 and do some stress testing. You'll know if your CPU is not functioning properly or not with that test.
I had these questions in general and have been waiting for this video, thanks.
Dude, thanks so much... I only have the 0x125 BIOS update. I updated the BIOS, but I never updated the MSI ME firmware...... I noticed the small font and downloads, but I didn't do that until I heard it in your video... 0x129 isn't available yet on my Mobo... thanks man!! I'll keep checking for the newest.....I'm running a 14700K... never experienced any issues... but it's a fresh build, less than 3 weeks old...
I'm worried that my chip is already fucked. I know they extended the warranty to 5 yrs basically, but I had ghost of tsushima cause a bsod once (only game or app that has done this)
Worried it's already fucked and have no idea how to tell if it's actually messed up....
One crash is kinda nothing, I wouldn't worry.
One crash could be anything, try not to worry bro. Hopefully you're all good 👍
The thing about this and what most UA-camrs have touched up on is that any possible damage is irreversible. BUT if there aren’t any more issues since the BSODs you mentioned, then there shouldn’t anything to worry about since there shouldn’t be anymore degrading
BSOD does not mean it is a CPU error, could be a lot of things in Windows!
Basically uninstall your gpu drivers and reinstall them 10 times, if nothing crashes it's probably good. Zipping/unzipping are workloads that are very sensitive to this + more than often decompressed files are checked against checksums
Can you test this with 13600k and 14600k
i5s already had lower voltages either way; bcs they re boosting way less than an i9; kinda same as i5s for the i7s; that s why they were at a lower risk
Yeah was wondering about the 13600k too, despite it being safer. Bought one recently, and I set a negative offset for voltage soon after ('cause summer ambient temps). Apart from a not so great single core score on cinebench, it's been working great, so I'm in no hurry to update. I'm sure someone will test the new microcode on them soon enough though.
A patch for a chip that is already permanently damaged, and also effectively beyond repair.
Intel is going to be writing volumes of refund and replace checks. Pre-built manufacturers such as Dell will also be doing the same. At a minimum imagine what those checks will be like for: the minimum of a replacement cpu and motherboard. If it’s an item like a pre build (given how many items are permenantly attached) it will essentially mean replacing the entire system
lmao are you hard of hearing? Jay literally explained the bios in his first statement. Go buy the embarrassing amd 9k cpus and st fu
Thank God you post some videos about motherboard manufacturer are making processor broken. I just recently bought i7-14700k and learned to undervolt and use the Intel default profile.
Would be really interesting to compare scores with release benchmarks (for the first "save the cpu" release I would have moved every knob down a lot)
The proof that Intel issues are manufacturing defects is the 13th gen cpus crashing now. It seems those CPUs are aging due to issues in the manufacturing process.
Let's see. I hope all issues are fixed and Intel learns to change