How can we create a positive outcome from this??? Let me know! --- 👉SPONSOR 👉 25% Coupon code for software using "TL25": biitt.ly/XQT7V Watch This Next! The future of GPUs is a disaster ua-cam.com/video/bc-lTgB8Ff8/v-deo.html 👉👉👉👉👉 THE COMPUTER 👈👈👈👈👈 (Links are Affiliate Links when possible) As a retail affiliate, I am paid a percentage of qualifying purchases at the expense of the retailer. 💎 CPU 13900k - amzn.to/4dG0eFE 💎 GPU 3090 - newegg.io/ncf0bf56a7 💎 Motherboard AQUA - amzn.to/4dkIjVg 💎 Memory 96GB - amzn.to/3Td3wZs 💎 PSU 1000w (thank god, given the UPS issue in this video) - amzn.to/3yCWsOC 💎 SSD 850X - amzn.to/4dkkhcX 💎 Watercooling - yeah, that's gonna be a lot to list. Comment and ill try to reply.
Your CPU was probably already degraded and too far gone hence why your voltages were also higher. You can't fix an already degraded CPU. You should RMA it... Period the end. This fix is to protect CPUs that aren't already degraded from destroying themselves. You even showed in an older video your CPU exhibiting symptoms of degradation.
Yeah, my understanding of the whole CPU degradation issue is that once it's apparent, the CPU is headed for that great e-waste heap in the sky: the degradation *can't* be made better and will only get worse, until the CPU eventually fails -- and that 0x129 can only (at best) _slow death's roll,_ as it were.
That doesn’t make ANY sense. If the cpu was degraded to that point that u are claiming…there is no way the cpu would be hitting those voltages without crashing. I’m afraid u don’t know what u are talking about, friend.
@@godnamedtay degraded CPUs require higher voltage to remain stable at a given frequency. Safest way to deal with a degraded CPU is to lower voltage to a safe level, then reduce frequency until it becomes stable.
@@CossackHD I would just throw it away if it’s that noticeable to where it’s no longer stable. I mean if u ran it into the ground, while most people have any idea what or how to do anything to prevent it. Either u under volt from the jump or if you’re lowering voltages to a noticeable level just maintain stability, that mf needs to be in the garbage. This is why I rma’d mine and tested the new bios, (it sucks) and reverted back to my most stable version and undervolted/ changed line load.
I've gone round and round trying to get them to handle an RMA. They've accepted, then said I had to pay shipping, then denied it, then accepted again, then said they can't get me a replacement, now they said they would refund me, but not till they have my original. All in all, ever fix they've offered, which is unacceptable really, leaves me without a system for an untold amount of weeks. What a MESS
@@Bareego that's the thing, they said they don't have one. Which makes me wonder, have they even solved the issue? Or are they holding stock now because so many are failing and this is a stop gap attempt
Omg so they don’t have a replacement lol I have a 14900k that been crashing in escape from Tarkov and was about to rma it But my voltage’s haven’t went over 1.55 in Tarkov since I changed the bios settings Still haven’t jumped to 129 yet
Intel's microcode update was meant to get rid of extremely high voltages (over 1.55V). Those are voltages that are very hard to pick up by software like HWInfo as it's very slow updating data, you need an oscilloscope for that. voltages in the 1.40s is normal and have been fine in the past, also for AMD
So we should just undervolt to be safe, if it proves degradation in warranty period, RMA it. ( Though i_d prefer the mobile PC supplier I bought it from, would offer a damn AMD alternative in the USA) Currently only offering it EVERYWHERE Else...sigh
@@lasseberg7317 Actually hardcore overclocking has a follow up video (Optimizing 0x129 & 14900K MSI Pro Z690) and was able to lower this 1.55v peak to around 1.44v while achieving slightly higher Cinebench scores in the process but the way he does it would require some homework on the user end especially if running a different CPU.
@@Azureskies01 what are you refering to regarding that it didn't really do anything? he has confirmed pleny of times that it did stop from going over 1.55V
So many people don’t understand the problem, how it was handled and what was even happening. And then you have some of them like this guy spreading that misinformation. The voltage issue was with the 1-2 core boosts, mainly eTVB, requesting voltages that were degrading the CPU and voltage spikes. HWinfo64 and other monitoring programs normally aren’t set to update fast enough or might not even pick the spikes up anyways. The microcode limited the VID to 1.55v, which is low enough to prevent degradation since it will only request that in loads where it is using the 1-2 core boost. If your CPU is already degraded then this could cause more stability issues and your V/F curve will be worse. That means it will use more voltage to maintain the same clock frequencies, lower the clock frequencies or become more unstable. Your CPU needs RMA’d if that is the case.
I'm buying an Intel CPU. What should do to prevent it from degrading. This whole situation is so annoyingly confusing. Once I update my bios should that fix the issue and keep my cpu healthy in a sense. And will it affect my cpu performance negatively? If so in what way?
@@JayUchiha17 you might see 100MHz lower TVB clocks, but that’s only in lighter loads and not even worth leaving enabled in my opinion. The latest bios revision for LGA1700 boards should fix it, but you can apply a slight undervolt if you want. That will help out slightly with lowering the core temperatures as well.
@rekareaper I thought long and hard about it and I'm really building a gaming pc. I canceled the intel i7 14th gen cpu and I'll return the motherboard, and I'm getting the ryzen 7 7800x3d. I shouldn't be spending all this money then troubleshooting something that everyone has a vague idea of what the solution is. I'm not buying a problem and then worrying about how I'm going to diagnose, fix, or return it. I'm also sure the resale value of these CPUs are going to go to shit years from now.
@@JayUchiha17 nothing wrong with that. The X3D CPUs are great for those who don’t want to tweak anything and just run them out of the box. As long as you use a bios with the 0x129 microcode, the issue is resolved. Just don’t use the auto OC features on the boards as those compound the issue and have never been that great anyways in my opinion. You’ll most likely be perfectly happy with either for one just gaming.
@@JayUchiha17 smart man. I was about to say you are crazy as hell if you buy intel. I’ve lost a 12900k and might be loosing my 14900k. Never again. I’m migrating workloads to MAC and I may build a new AMD system for gaming etc with the rest of my parts.
0x129 is extremely stable on my 13900k box. Z690 formula. CPU is also requesting lower voltages than before. Also, before I couldn’t run my memory at 7200 MHz stable, the system would inevitably blue screen. I had to step down to 7000 MHz. And I was having all sorts of random crashes in only certain games like Forza Horizon. With 0x129, none of that is a problem any more. 7200 is stable. I even applied a slight overclock, a per core turbo ratio overclock. boosting to 5.9 on 1 core, 5.8 on two cores, 5.6 on 3, and 5.5 on 4 or more cores. I could probably push the system even further. But with this dialed in VIDs max out at 1.35V and average 1.24V. For me, 0x129 is what my system needed and it is what I shall use until it’s time to sunset my system for a newer Lake. Or possibly an x3d.
For me the new microcode actually made my ram unstable. I had to adjust my timings to make it work. But yes the cpu voltage is better now, before my cpu needed 1.31v for 5.6ghz all cores, now i needs 1.275v
@@angelskeet1532 go to bios and set a vcore voltage. Dont forgett to lock all cores to same speed. The two boost cores dont add any performance increase in gaming or multitasking. Then you can try something like 1.35v and lower the voltage until you find the lowest possible stable voltage.
not on msi boards. unfortunately I bought a z790-a board from them, and I have no option to limit voltages. needless to say, I will never buy an msi board again. not only because of that, but also because of the endless amount of bugs with their bios.
@@Stormpriest only advice I can give u is do not update the bios unless you are 100% sure the version ur about to use is clean of bugs. my experience was.. insane to say the least.
@@jacobgaysawyer337 bro wtf are u talking about. First of all, yes u can, secondly, have u ever heard of xtu? lol this is nonsensical to talk negatively about Msi mobo’s because of ur incompetence. Jesus Christ.
The patch was to fix voltage spikes that you cannot see with the motherboard sensing software. You need an oscilloscope probed to the voltage rails of the CPU to see it. In many cases the voltage pre patch was going above 1.6V for a few microseconds, but it was doing it very frequently usually between a high power state to idle where the CPU was requesting the wrong voltage, this is made worse in some cases depending on the power delivery components on the motherboard. Increasing the voltage overall might help with stability if it was unstable to begin with. Will it decrease the life of the CPU, maybe. 1.5V does sound rather high, AMD CPU's are 1.35V to 1.4v, but I assume it's entirely on the process used by each company (Intel vs TSMC) to what each can handle. The oxidisation issue was limited to the 13th gen on batches between a certain date range (that Intel have no disclosed). The damage to the CPU is caused by a process called electromigration, you can read about that on Wikipedia or other places.
Highly appreciate your very relevant note on electromigration. For others who see this, a picture is a thousand words, and you can see the effect on the wikipedia page.
Intel users dealing with instability and micro codes that increase power draw, while AMD users downloaded 10% extra performance for free! Edit: Guarantee that Intel chips will have more stability and degradation issues after this "fix", and will have to come out with a new fix for the previous fix. RMA nightmare.
"while AMD users downloaded 10% extra performance for free!" Let me fix that: People freshly installing a preview-version of Win11 get the performance-increase of disabled memory security.
I had memory integrity disabled from the start. Personally, in two different cpu limited games, I’ve seen 7-15% increase in performance with a 7600x. And you don’t need the preview version, there’s an update for everyone to download. 👍 If you watch the latest HWU you’ll see just how wrong you are about the memory integrity option.
The bios change brought some changes to the bios, but board partners changed other things at the same time. Some changed the AC and DC Load line Ohm settings at the same time and some did not, leading to a very messy situation, because the change changes the CPU voltage.
I love Intel, but 2 of my 13900ks failed. That was with undervolt of 1.28v, cores pinned to 5.5 ghz and 4.3 ghz (P-cores/E-cores), and no boost to 5.8 ghz. Both failed at 6 months mark, at same exact way. Both were brand new sealed in the box. I moved over to AMD, because I could no longer diagnose the issues
What happened when they failed? I have been running fine with my 13900ks for 2 years but since last month im starting to get kernal 41 error. Hard reset. Not sure if its the new bios or if my cpu have now failed to work properly. Its passing all stress test for cpu, gpu and ram but still crashes randomly.
Dumped my 13900k and 13600k. I’m tired of getting no fix of the issue. Got my 7950x3D, zero issue now, and I can run with eco mode, so much more energy efficient.
One detail has stood out to me since the 0x129 microcode release before I heard anyone say anything: Intel didn't say it was a fix, they just said it "addressed" voltages. Therefore, it's not a fix but a method of keeping performance as advertised and merely slowing degradation.
well, degradation is not reversible and, at some point, inevitable. The microcode was mismanaging voltages, making the processors degrade faster than advertised. If, and that is a really big if, the new microcode keeps degradation times in specs, then it is the only fix they could ever provide.
@@Niyucuatro I look at it a little differently. It's the only fix they could provide that would keep them from being sued performance-wise before the warranty runs out which allows the cpu's to degrade slower than before the update while not actually fixing the problem of degradation. Intel has to define what the specs (voltages) are supposed to be (they haven't, so there are no specs where it really counts), keeping in mind the fact that if any microcode update slows the cpu down enough to make it be below the advertised speeds it could and probably would be considered false advertising performance-wise with impending lawsuits that Intel doesn't seem to be able to afford considering they didn't do a recall of the affected raptor lake cpu's. So to summarize that, I suspect that they aren't willing to say "This microcode is a definite fix" (for all cpu's that haven't degraded) because they can't lower the voltage enough because it would hurt performance enough to be sued. They're walking a very thin line here and I suspect that they need to lower the performance even more to make them stop degrading except they can't.
That's not really how CPUs work. I can upload a video running my 13700k at 1.8v if you like. Worst case scenario instead of the CPU lasting 100 years, now its only going to last 95 years. You will not notice it.
Since the moment i got my 13600k, i noticed high temps. Even with a nice water cooler i was getting around 100ºC. That was unacceptable for me. I reduced voltage and performance core ratio to like -0.150v and 47x, directly from Intel xtreme tunning software. I lost performance on my benchmarks, but now my cpu gets 78ºC max. For a long life cpu, that works, but it's too bad to know that my cpu is limited. I did this a year ago btw.
I haven't seen my 13600K go above 60 degrees during heavy normal use like cpu demanding games. It goes up to 80 degrees briefly during Intel Tuning benchmarking. Yours was probably defective out of the box.
Bro it was 2+ weeks from the day I started my RMA for a new 14900k. The cross shipping form I sent back, Intel tried to tell me there was an issue with my info and management would get back to me. 2 different promised call backs that never happened later, and never being able to since reach anyone on the phone with my ticket number, even after waiting over an hour on hold multiple times…via the Intel chat on their website, I threatened to and showed them the documentation for 1 of the class action law suits I was going to sign up for, I then received a phone call within 15 minutes and my replacement cpu was shipped to me with next day air UPS shipping….lol this is a true story.
@@_Mattz1nho I had to give them my credit card and they charged $450 and held it hostage until they received my old cpu. Although, because I complained so much about how much they suck and how long they ignored me, they ended up covering the $25 shipping charge lol
@@_Mattz1nho they charged my credit card for the price of the cpu until they received the defective one. Trust me, u WILL need to follow up about getting ur money back or else it will take up to 30 days.
Best to use modified settings with 0x129 BIOS. I did analysis on Asus ProArt Z790 with 14700K: ua-cam.com/video/J2k9_dqKlnw/v-deo.htmlsi=f8MFkxBFxE7U625c at 17:00 in video is where I show comparisons. Bottom line is (and probably for other CPUs as well) with a few settings you lower temp/power/vcore substantially, with almost no performance hit, and there is no thermal throttling. With 0x129 BIOS as is there is thermal throttling, vcore still goes up to 1.5 something. Modified settings keep things more under control which is just another means to increase likelihood of stability/longevity.
Had a DDR5 4 DIMM configuration (yes I know its not recommended). Was working fine even with XMP at 6000 Mhz. I then updated my BIOS with the new 0x129 microcode, got instability, bluescreen, crashes, FPS drops etc. I removed 2 RAM sticks and disabled XMP. Changed to "Performance" profile in BIOS and it is stable again with some rare crashes etc. Sooo frustrated, so I ordered an AMD CPU now.
Hello, if I just bought a new PC that has a 1400KF, should I first test it to see if it has the overheating issue before updating the BIOS on my motherboard? or should I just update it right away?
Appreciate the effort you took into making this video but it is extremely badly researched and your voltage numbers are not correctly measured, nor are they relevant to what the actual problem is. The voltage spikes seen in the older microcode were in excess of 1.6v, and they were not seen in the middle of heavy, multithreaded workloads, they were seen during transients of applying/releasing loads. This is in line with some of intel's communications on the matter. As others mentioned, this fix will not help CPUs that have already been damaged, it will only stop (further) damage to CPUs that are still functional within the operating margins.
I was afraid this was a problem, I have a 13900k and I bought a used RTX4080 about 6 months ago. I thought it was the GPU. The system just randomly boots. The windows event viewer shows nothing. I did the microcode update and it's still happening, at this point I don't know what to do?
It's funny how there are double standars. Some reviewers don't use PBO on AMD, saying that it's not the default profile, and therefore limit its performance. But when it comes to Intel, they are willing to try out 10 different profiles, under volting and using massive water coolers, while saying that's normal.
@@abaj006 I had to do so much settings In the bios in an attempt to make the intel chip stable (and it still wouldn't be) or become less hot, that I'm used to it. When I switched to AMD I heard about PBO, which made the chip run cooler, with more performance. I was sold immediately. Easiest "overclock?" Ever. And I always changed these settings conservatively. I'm nit pushing the limit, all I want is for it to run a little bit cooler, so I don't have to replace it as soon.
In this video I just saw a lot of misconceptions on how modern CPU operates: 1. what you refer to as high temperatures is absolutely within spec; 2. the UPS beeping doesn't have anything to do with CPU issues, if anything your UPS is underspecced; 3. Vcore is not what the CPU gets but also what the CPU requests, there are multiple independent power lines going to the chip; 4. your high voltages are also within spec as stated by Intel, and while it's not normal to have them in idle it's absolutely expected to reach them under load; 5. when you say the CPU is throttling it's actually not as it's still over the baseline clock, it just limited the turbo-boost and this is absolutely normal. If there was a problem with high voltages under certain conditions that got corrected with a microcode update, it doesn't mean this problem was present under all circumstances so it's a wrong expectation to see your voltages decrease, you might have never experienced the high voltage issue that Intel fixed.
i brought my 15-13600k in january 2023 - havnt updated the bios - do i update the bio even if i have had 0 instability issue - it does run hot yes but no crashes at all
I recently updated to the new microcode. Im running a 13900KF on a Strix Z690-G. Intel Extreme Profile, Enforced Power Limits, MCE off and Intel Fail Safe VID behavior gets me 35K on cinebench R23 and 1.50v max VID. Otherhand, Asus Advanced Profile, Disabled Power Limits, MCE on and Auto VID behavior gets me 40K in Cinebench R23 and 1.42v max VID
I bought a 13600 k last week, its installed on z790 asrock pg lightning and i have aio id-cooling 240. I have no issues with default bios, temps are on idle 26-28 degrees, and on gaming 40-50 degrees(summer) I cant complain for now.
the OS also updates microcode via updates, those are always applying the newest, coming from the bios or the OS, you should check always the current microcode of the running CPU while testing versions
Hi @TechLens - can You write if betwen all that problems in cpu stability did You notice any problems with booting system? For example: explorer.exe is nit automaticly starting after log in to system or even worst the logonui is not showing after initial system boot? Any common error logs that are repeating when cpu crash?
Many years ago, when we were in the 32nm range, 1.5v was the voltage where CPUs started eating themselves. I can't imagine that shrinking geometries has raised safe voltages. I would rather operate in the 1.3v area.
A lot of the crashing was due to motherboards undervolting due to load line calibration issues. Where the microcode issue was causing transient spiked above 1.55v which appears to be the danger zone. These are two separate issues. Undervolting average voltages and overvolting in transient spikes. Hence the intel fix is supposed to both raise average voltages to improve stability and reduce the peaks of the transient spikes so they don't go above 1.55v. Addressing both the overvolting and undervolting. This appears what you are observing. Raising voltage and reducing power limit is of course going to impact performance. But you can manually tweak this with an undervolt and recover the performance.
TechLens : the safe continuous voltage for the 12-th, 13-th and 14-th GEN Intel cpu-s is 1,30-1,35 MAX. All voltages above that are dangerous the moment that they become more than just nanosecond peak ones. The problem arises when higher values of current are pumped into the cpu and in correlation with the 1,30+ Voltages the CPU starts to draw abnormal amounts of PWR. The damage occurs at that point but it also occurs even if little current is pumped into the cpu but the voltages are continuously higher than 1,30 V.The degradation is a result of certain portions/clusters of tranzistors literally being melted together.
thank you for this video and sharing your experience, just one question, if VIDS voltages are what the CPU are requesting why we focus so much on monitor them instead of monitor VCore which is what was actually delivering? what I am missing here
I'm pretty sure intel is just trying to cover it up until the CPUs are out of warranty range because a recall would make them look REALLY bad. That would explain why the voltages are boosted. The performance is already degraded so if all they have to do is boost the voltages then they'll do that knowing most people will blindly apply the firmware updates. This is a large corporation which is suffering pretty big losses and has a history of shady business dealings. I don't know what's actually going on behind the scenes or anything just speculating with what's publicly available.
With software-readouts you can NOT check for voltage-spikes. So just from that alone your "testing" is worthless. And the 0x129 is just there to enforce the 1.55V VID-request limit. The UEFI actually has a lot more protection-features but most motherboard-vendors had decided to disable them all.
1.55V is still far too high. Anyone experienced in overclocking will tell you that is a very dangerous voltage that WILL brick your chip eventually unless you have the goldenest of golden sample silicon.
@@asdf_asdf948 "Anyone experienced in overclocking will tell you that" Your assertions are wrong. 1.55V is really high, however the thing that will cause damage and why high voltages are avoided are the spikes that go beyond that. If you lock your CPU to say 5GHz allcore at 1.55 V (As people did and still do for that) then you will not have problems aside from the CPU constantly wasting power. And the high voltages only get applied when you get 1-2 cores to full load - and then the problem is still when the load actually ends and you get overshoot.
Just one question as owner of an I9-13900k myself who has experienced random game crashes and even reboots in the past myself though it always was rathre random. Did you try to assign a limited set of CPU cores, like only the P-cores to a game and remove the other cores, like the E-cores, from the game and let it do other stuff instead? Did that help reduce the power consumption? From what I noticed, this seems to make games more stable like in Star Citizen, which used to be very CPU heavy before their graphic updates, Red Dead Redemption 2 or Hogwarts Legacy.
I'm having a tough time optimizing my I-914-900K. I just got a replacement from Intel as I have had nothing but issues using the ASUS motherboard that automatically overclocks dark hero x5790. Even turning that off completely doesn't stop my CPU from thermal throttling. It never overheats but it thermal throttles every benchmark that is CPU intensive. Not sure what I should do if I should undervolt it or leave it at stock or overclock it
I had to RMA my 13900k to Intel I was upgraded to a 14900k due to supply. I shipped via UPS and had my CPU back in 7 days . I bought a contact frame and Kryonaut Extreme and haven't had any issues after bios update.
Hi, do you know of any work-arounds for laptop users who have an intel 13th and 14th gen CPUs? Let's say if their laptop doesn't have a problem and wants to at least avoid the problem that others may have on Intel's ongoing issue.
I have recently bought 13900HX Laptop. Its been a month now that I am using it. Played many games since then especially Black Myth Wukong. I don't see any crashes on it. Hopefully laptops are not really affected.
@@blenderesque708 did you do anything with it, such as update the laptop's BIOS, download Intel's microcode, update the laptop's driver from the manufacturer's website?
if you need more voltage to fix crash, it's not a good sign, how long will it last that cpu? and how much power draw required for that shader compilation? 400watt?.....can you do a video with undervolting? because doesn't seems a fix but just more voltage to the ones who has a degraded cpu
yes that's right, it did also add a 1.55v VID limit when using the Intel pre-sets (that should be the MB boot default now). Also set 253W for the PL1 and PL2 400A for ICCMax (it actually did this part in the least update too) previously they were often set at unlimited!
Idk how this happens.. ive had 13900k since launch and have done the bios update for micro code fix. No issues at all. (Pre and Post Update). Temps dont sky rocket either. Or voltage spikes. (I dont overclock)
I recently updated my bios to the new microcode, for some reason MSI took there sweet time releasing it for my motherboard. The big thing I noticed was the clock speed was way down. I have a i5 13600k which is supposed to be 5.1GHz. Now the best it can do is 3.5GHz
@@fofal Did that and still acts off. I'm not getting the stutters I was getting before. Doing work related things, or having multiple windows open I was getting a lot of stutters and lag. But now the speed and voltages are all over the place. Default settings don't change it. Undervolting I can get it to stay closer to 4.4GHz, and using Intel Extreme Tuning's automatic overclock I can get it to 5.1GHz. But again the speed jumps all over the place. Maybe the cpu is defective and it's in it's early stages, I have no idea. Never had an Intel cpu act like that and i've been using them since 8th gen
@@spentcasing3990Is it overheating? What is your long duration boost (PL1)? I assume previous BIOS' undevolt caused your chip to run at its edge of stability hence the stutters
When you take a deeper drive under the hood... Exactly how are you taking this deeper dive under the hood and does it involve something more than software that monitors voltages and cinebench?
yes return it and get a 12th gen processor if you want to stick with Intel. It's the latest generation that doesn't suffer from these problems. Why do you even think the guy on eBay was seeling it at this point in time? Also wtf does refurbished CPU even mean? You can't refurbish a CPU as far as I know. If you can't go 12th gen Intel (which btw has the same LGA 1700 architecture as 13th and 14th gens so it fits on the motherboard for those gens as well), go AMD
Recently replaced my CPU and mobo, went from Intel to AMD, and no regrets. What I would also caution against is using Intel's RAID, as that made the switch more difficult than I want in the future. Just an FYI.
Im hesitant to upgrade. My two month old new pc runs great and plus it’s an i5 14400F and temps and volts never get out of line. I haven’t heard if the i5’s are affected
Though I have experienced no issues (on ASRock Z690 BIOS 13.05, now two years old), my 13700K remains in stasis while reviewers get this fully sorted out. I know that at some that point I'll have to fire it back up, apply the latest BIOS/microcode, and see what happens.
i had no prob with gigabytes new bios microcode update. and perhaps you look at wrong type of voltage elevation. it requests specific voltages at specific times
TLDNR YES - update to Bios with 0x129 microcode. Then if you are comfortable with changing Bios configuration select the Performance intel setting if you want to be a little more conservative (it basically lowers your power form 400A to 307A) or stick with the Extreme setting (default) if you want best performance. It might run a little hot but changing anything else to optimise it needs some knowledge, check out some other streamers too on the subject. (hint - It will involve AC/DC Load line changes and CPU negative voltage offset ( undervolting))
I’ve noticed a few things with the new bios for my i9 13900k on my strix z790-A. My temps are higher, games that were butter smooth now crash and have stuttering. My temps were very good under 45C during heavy gaming. Now playing the same games I’m hitting 60C+. Games that wouldn’t even launch before (vram error) now launch properly but have Intermittent crashing to desktop. I spent WAY too much money for this to be happening
It's current that causes physical damage. You gain stability by increasing voltage and decreasing current (among other things). To decrease 'average' current you can decrease frequency. Because of the reduction in frequency, less work is done per time; you must run benchmarks to measure work rate.
I have always been on the safe side, all the time my multi core option on asus was turned off. Lately I have been observing strange things, performance drop after undervolt, different results, on the OCCT application at 100% load I do not reach speeds above 4.8 MHz Processor + Ram 5.2, on geekbenchmark it is solid 5.5, with turbo 5.8 We are talking about 13900K On cinebench R23 on multicore I cannot reach 37K mostly around 36600, 36800 A strange smell also appeared a week ago, it was for a few minutes but it's strange It's hard to say even after tests what condition the processor is in, and the motherboard is also unknown in what condition it is after such loads.
Well the intel info said "Address high voltage" now "Lower high voltage". So maybe they did some tests and decided they'd want to keep the voltages up so they wouldn't tank so much in benchmarks.
I'm not a Pro in this, just a newbie. Fearing supply chain wars from China, and having an aging but working original 14 year old i7 860 HP Pavilion computer, I built myself a gaming PC with MSI's Carbon WiFi Z790, i9 13900K, 4090 GPU (both liquid cooled) and GSkill DDR5 5600 32GB x 4 (128GB) RAM. Before I updated the microcode my XMP profile showed 5400MHz RAM speed but 5.3GHz CPU speed with the CPU in "Game Boost" mode. Now, after the x129 update, the CPU now runs at 5.9GHz (and the RAM runs at the same speed, i.e. 5400MHz). So the question is: which is better for stability and performance: higher or lower CPU clock speed? I checked my voltages compared to my previous setup (from last year's BIOS), but they haven't changed much (maybe 1 or 2mV). I haven't benchmarked my computer yet.
The performance on the latest i7 and it’s was so impressive that it’s such a shame this whole malarkey has happened. Mistakes happen, but to push it out to the public when you know there are such issues, and then when people RMA the product for valid issues just to shirk responsibility, and then not to fix the issues from happening, and then to not be transparent at all about the issues is simply malice, rather than incompetence. It’s so depressing to see these great American companies like Boeing and Intel decide to profiteer instead of innovating and ruining these companies chasing short term profits.
It's a red flag to read voltage upwards of 1.45v but 1.72v? I don't think I'd ever buy another intel chip again, I wouldn't care how much they tell us that this is normal behavior. 1.72v.... unreal!!
Yes, but you didn't say for which specific cpu because you don't just lower voltage, you have to lock core frequency too to keep it from boosting too high at stock frequencies requiring more voltage just to not crash.
In his case I assume it was some crazy AC LL shipped with the new bios, like 1,7 mOhm (it happened to me after updating my biose I got 1,4V instead of previous 1,3V before any undervolt). Typical AC LL should be between 0,01 - 1 mOhm depending on your LLC (the steeper the curve, the more AC LL you need to compensate, but it should never go past 1,1 mOhm).
Seeing that you not only use a UPS with your PC, but you have it affixed to your desk /respect. I've been using them for about twelve years now. Initially out of necessity due to living in an area that had constant brownouts, but then just for peace of mind. My current residence doesn't have constant brownouts, but it does have power fluctuations every few months and at least two times a year the power goes out completely for about eight hours.
This has been on the stove for a few years now. Ever since the rise of Ryzen, Intel has pushed crazy power into an architecture that would eventually fail. Which it has. Their “P” core / “E” core experiment has been a huge failure. There was no possibility of keeping up to AMD’s core count on that architecture and I’m sure there were some Intel engineers who voiced their concerns………but were ignored. I’m not sure I will build Intel rigs for a while moving forward. There is no trust remaining.
3:46 - most likely just combined power consumption of your cpu + gpu with unlimited fps. For me, even shader compilation takes 20% less power than CB R 23 and 35% less power than AVX2 benchmark like OCCT Extreme.
I solved the problem by capping the clock of my 14900K to 5.4 Ghz. There is no other way around it. Wattage = Voltage * Amperage. You can't have 6Ghz on this architecture, period.
Waste of money really. Really bad a look for intel. A cpu that can do 5.7 all core and more in some cases has to be capped at 5.4 just to work. RMA it brother
There are other ways round this. 0x129 has stopped the damaging/ crashing spikes (1.6v+) but still set voltage at 1.55V which is too hight in my opinion and could still degrade the CPU over time, but has meant less crashing. So alternative to locked frequency is you can go into the bios and set a max voltage yourself, (say 1.4v) then under volt the CPU, -0.100mv offset in my case, your CPU might be able to take less or more, trial and error. That gives you full range of frequency’s (inc 6000mhz boost) and the system never goes over 1.4v. (oh and thermals are better too)
Just check your AC load line. MSI for example is pumping 1,7 mOhm for some boards with the new bios. It means +170 mV per 100 A minus LLC impendance. Max LLC impendance is 1,1 mOhm for most boards. If you are using LLC auto (steepest curve) your AC LL should be within 0,8 - 1,1 range.
No, all chips age with use, its just that lower SKU Bins running with bad bios settings will show problems (degrading) first. It seems to me that what Intel are trying to do is get rid of the bad bios setting problem and a microcode update to stabilise the problem so your chip lasts at least 5yrs (warranty) If you a home user you should get 10yrs out of your CPU, unless your running a server farm or mining 😁 Hey you don't run your Car with the foot to the floor do you 😜
The retailers in system integrators are doing the right thing by extending the warranties. They are going to be hurt the worst. They won't recoup the time and labor, shipping cost to replace all these over the years. The margins and profits can't balance against this future level of repair.
Is there any fix to prevent overclocking and crashing with certain software and games, using a computer with an Intel chipset? I haven't noticed one yet that's actually legit, so I'm almost at the point of giving up and buying a new mini PC with AMD (If I can find a good one that is).
Meanwhile, my 12 years old Core i7 3770K runs @ 4,5 GHz with a Vcore of only 1,254 Volts. And I can still play Cyperpunk 2077 @ 1080p max settings @ 85~88 fps 😂 Ivy Bridge FTW, mates 😁🥰😂😂
When I got my 5600 it would run at 1.5V+ default on my b350 f-strix and wouldn't hit 4.6 GHz singlecore pbo only, reaching ~4600 in cpuz. After me overclocking, I'm now running 4650mhz all core @ 1.2 V + LLC goes up to 1.263 V hitting around 5100 in cpuz. Still lower temps on the vrm and CPU. I feel like they are all trying to fry our hardware, so we can buy new...
bios 0x12B solve all... Got more perf, stability, better temp... install asus suite, launch Ai OC, keep scale at 100, and when you reboot, ai suite will update correctly your bios. After you can disable it. I have disable TVB for max Perf. Got stability, perf (6ghz) and good temp.
The MAIN problem is that intel is calling this a FIX but this is not a fix. If your bread is already cook it wont uncooked it. This will only prevent your current uncooked bread to not burn as fast.
They use "fix" as a word because if they didn't, no one would download it period, even if it does fix the nasty voltage spikes above 1.55V like how it does here. If your CPU was already kicking the bucket, this won't fix it. Nothing can fix an already degraded, about to die CPU ya know.
Easy way to not hit the high vid request on the new baseline and microcode update is say lock your two preferred cores to the rest like a 14900k is 57, then up your IA VR to 1400-1420, that should give you just enough voltage to have all your p cores to hit 5.7ghz since they will have the power to do so your VID will top out at 1.36X-1.38Xv, then the Vcore will also stay around 1.35V. It could take less depends on the quality of your chip. My P and E cores are not that great, so need a little more for my chip specifically. Do not want to do all of that you can just disable your E cores just expect your load times to take a little longer.
I want to build a new PC, but am totally confused, don't know what to do. The GPU (NVIDIA) problems with the connectors, and Intel with the problems. I have had the Intel CPU and NVIDIA GPUs since the early 2000s. I want to keep it that way, I know nothing at all about AMD, and don't really want to, I believe :)
Nvidia connectors are fine, i've had a 4090 from the day it released and all i did was make sure the connector was fully clicked into place.. also since about 9 months ago Nvidia made the connector a little shorter so they have less wriggle room for people to not plug in correctly.
@@_Mattz1nho Anecdotal evidence is hardly assuring that a statistical problem doesn't exist. Even if you don't know anything about how many of these have issues, just looking at the new connector compared to the previous connector, it's remarkably undersized for the increased power draw.
My Coffee Lake CPU doesn't suffer from any of these issues. 700mhz all core overclock constant at just 1v to the CPU, under load my CPU barley pulls 30 watts CPU Package Power as a whole. It still bangs out 90 fps at 1440p gaming in new titles. Sure it's old now but it works so why replace? Don't get e wrong i am in the market for a newer CPU MOBO combo, but not 13th or 14th gen intel and i do not like the fact that X3D chips do not like to be pushed beyond stock which is something i have always enjoyed getting more from. As the years passed the stock 4ghz was not enough so here we are with a good overclock and it cost me nothing to extend the rigs life an extra 3 years or so.
Software is NOT an issue. I tried windows 10/11/Windows Server 2022+2025Preview & (5+ versions of Linux with different base) of which most crashed or froze during install. I tried new SSD and different RAM. I tried with NO GPU and with old 1070 GPU. At the moment I have to run CPU at x50 multiplier for it to be stable. All of the above I tried while rotating hardware around and yes DDU was done in safe mode. My place of purchase refuses to do RMA because it passed their 24h street test. My crashes are with normal work, office stuff which is about 95% of what I use the 14900k for. I Work in IT and can do all this stuff while sleeping. That's how easy it all is. BIOS is up to date
When you say "the values are something they check" you assume that the data is ever exposed to humans. There is no reason to believe that anyone at Intel has been looking at the values or their range. If Intel failed to monitor the values applied by automation and generated by formula, the voltages could have crept over time. Without a defined upper maximum, unlimited drift could have occurred over the years. Process changes over generations drift beyond the initial design spec for the code which optimizes and one day you discover that a blight has been creeping into the product for years. One of the dangers of extreme organizational complexity is that knowledge becomes compartmentalized and quickly lost. The old process may look like it works for the new product but you can only tune for issues which you either foresee or detect. If you hid a variable in the process which drifted over time , you are blind to it.
I have used AMD for the longest time and 5 months ago I built a computer that my company spent 5 grand on and I decided to go with Intel because of supposedly raving reviews. What did I get in return? constant crashes, weirdly slow computer almost all the time, even very light programs (like vcarve or illustrator). This is ridiculous. I should have went with Ryzen series or a threadripper. I really screwed my self here.
Intel is still working on microcode fixes. RMA your 13th or 14th gen processor for a full refund and buy a 12th gen. the 0x129 microcode doesn't fix the root cause. Perhaps the next microcode version will fix the root cause, but at this point just request a refund.
With these series of chips, we just need a do-over, this is such a mess. Intel is just stalling to avoid being sued. As long as they can say that they are doing something, they stay out of court. It seems it isn’t a matter of will these particular chips die, it’s when and will you be the in lucky person that will be left holding the bag if it’s after your warranty. Monitoring voltage is great advice after finding out how different each individual chip is. It is truly great that the Information Age has provided media outlets the ability to uncover what they use to hide. Great job here and outstanding coverage from a different prospective.
You seem to be assuming only one scenario is possible - either more voltage is being supplied or less. But I think what's actually happening is that the CPU is getting both more AND less voltage, as needed. What do I mean by that? I think a cap has been placed on the maximum amount of voltage the CPU can request, because super-high spikes were causing degradation, AND in parallel these CPUs are now being fed more overall voltage (minus the big spikes) to ensure stability. And by doing that, Intel is theoretically preventing degradation while also increasing overall system stability. Will it work long-term? Only time will tell. But for now at least, Intel believes 1.55v and under is safe - safe enough to get CPUs past their newly extended warranty. Personally, on my own 14900k, just to be extra safe, I've installed the new bios with the new microcode AND I've capped all of my P cores at 57x and applied an undervolt.
I am an Intel fanboi. That being said, any firmware or software-based solution, will not "fix" a problematic manufacturing flaw. At best, firmware may moderate performance issues in order to avoid using them. As for the overvolts, well to allow for performance that is at parity with the expected norms, it seems logical that decreasing efficiency of some paths would require some level of compensation by increasing efficiency of others (I would expect). To be clear, I am only relying on logic, I have no schematics of functionality nor an understanding of what is being performed by these patches. So, grain of salt. 30 years in the industry is all I have to work with. Without a team of engineers, it's hard to speculate. If I had to guess, no firmware patch could "fix" a manufacturing flaw. the best we could expect is to compensate for these flaws. Interesting perspective though. thanks.
RMA your 14900k if you can't undervolt it anymore without crashing. Once you have your replacement cpu, just sell it, switch to amd or just downgrade if you really need intel quicksync/performance features.
How can we create a positive outcome from this??? Let me know! --- 👉SPONSOR 👉 25% Coupon code for software using "TL25": biitt.ly/XQT7V
Watch This Next! The future of GPUs is a disaster ua-cam.com/video/bc-lTgB8Ff8/v-deo.html
👉👉👉👉👉 THE COMPUTER 👈👈👈👈👈
(Links are Affiliate Links when possible)
As a retail affiliate, I am paid a percentage of qualifying purchases at the expense of the retailer.
💎 CPU 13900k - amzn.to/4dG0eFE
💎 GPU 3090 - newegg.io/ncf0bf56a7
💎 Motherboard AQUA - amzn.to/4dkIjVg
💎 Memory 96GB - amzn.to/3Td3wZs
💎 PSU 1000w (thank god, given the UPS issue in this video) - amzn.to/3yCWsOC
💎 SSD 850X - amzn.to/4dkkhcX
💎 Watercooling - yeah, that's gonna be a lot to list. Comment and ill try to reply.
What UPS is that you are using? It's caught my interest.
having a video analyzing a tech yes city video is the blind leading the blind
Your CPU was probably already degraded and too far gone hence why your voltages were also higher. You can't fix an already degraded CPU. You should RMA it... Period the end. This fix is to protect CPUs that aren't already degraded from destroying themselves. You even showed in an older video your CPU exhibiting symptoms of degradation.
I wonder if CP77 crash caused the BIOS to reevaluate voltage/frequency table and increase voltage to keep the CPU stable.
Yeah, my understanding of the whole CPU degradation issue is that once it's apparent, the CPU is headed for that great e-waste heap in the sky: the degradation *can't* be made better and will only get worse, until the CPU eventually fails -- and that 0x129 can only (at best) _slow death's roll,_ as it were.
That doesn’t make ANY sense. If the cpu was degraded to that point that u are claiming…there is no way the cpu would be hitting those voltages without crashing. I’m afraid u don’t know what u are talking about, friend.
@@godnamedtay degraded CPUs require higher voltage to remain stable at a given frequency. Safest way to deal with a degraded CPU is to lower voltage to a safe level, then reduce frequency until it becomes stable.
@@CossackHD I would just throw it away if it’s that noticeable to where it’s no longer stable. I mean if u ran it into the ground, while most people have any idea what or how to do anything to prevent it. Either u under volt from the jump or if you’re lowering voltages to a noticeable level just maintain stability, that mf needs to be in the garbage. This is why I rma’d mine and tested the new bios, (it sucks) and reverted back to my most stable version and undervolted/ changed line load.
I've gone round and round trying to get them to handle an RMA.
They've accepted, then said I had to pay shipping, then denied it, then accepted again, then said they can't get me a replacement, now they said they would refund me, but not till they have my original.
All in all, ever fix they've offered, which is unacceptable really, leaves me without a system for an untold amount of weeks.
What a MESS
By now they should send out a replacement with a postage paid sticker in the box for returning your faulty CPU.
@@Bareego that's the thing, they said they don't have one.
Which makes me wonder, have they even solved the issue? Or are they holding stock now because so many are failing and this is a stop gap attempt
Omg so they don’t have a replacement lol
I have a 14900k that been crashing in escape from Tarkov and was about to rma it
But my voltage’s haven’t went over 1.55 in Tarkov since I changed the bios settings
Still haven’t jumped to 129 yet
Intel is garbage in multiple ways.
I just received my new cpu a free upgrade to 14900k from 13900k it took 2 weeks++ but finally received it today
Intel's microcode update was meant to get rid of extremely high voltages (over 1.55V). Those are voltages that are very hard to pick up by software like HWInfo as it's very slow updating data, you need an oscilloscope for that.
voltages in the 1.40s is normal and have been fine in the past, also for AMD
so true. "actually hardcore overclocking" covered this. 1.55V is still verry high though.
So we should just undervolt to be safe, if it proves degradation in warranty period, RMA it. ( Though i_d prefer the mobile PC supplier I bought it from, would offer a damn AMD alternative in the USA) Currently only offering it EVERYWHERE Else...sigh
@@lasseberg7317 Actually hardcore overclocking has a follow up video (Optimizing 0x129 & 14900K MSI Pro Z690) and was able to lower this 1.55v peak to around 1.44v while achieving slightly higher Cinebench scores in the process but the way he does it would require some homework on the user end especially if running a different CPU.
expect Actually Hardcore Overclocking found that it didn't really do anything.
Also no, anything over 1.3v for intel is killing chips.
@@Azureskies01 what are you refering to regarding that it didn't really do anything? he has confirmed pleny of times that it did stop from going over 1.55V
So many people don’t understand the problem, how it was handled and what was even happening. And then you have some of them like this guy spreading that misinformation.
The voltage issue was with the 1-2 core boosts, mainly eTVB, requesting voltages that were degrading the CPU and voltage spikes. HWinfo64 and other monitoring programs normally aren’t set to update fast enough or might not even pick the spikes up anyways. The microcode limited the VID to 1.55v, which is low enough to prevent degradation since it will only request that in loads where it is using the 1-2 core boost. If your CPU is already degraded then this could cause more stability issues and your V/F curve will be worse. That means it will use more voltage to maintain the same clock frequencies, lower the clock frequencies or become more unstable. Your CPU needs RMA’d if that is the case.
I'm buying an Intel CPU. What should do to prevent it from degrading. This whole situation is so annoyingly confusing. Once I update my bios should that fix the issue and keep my cpu healthy in a sense. And will it affect my cpu performance negatively? If so in what way?
@@JayUchiha17 you might see 100MHz lower TVB clocks, but that’s only in lighter loads and not even worth leaving enabled in my opinion.
The latest bios revision for LGA1700 boards should fix it, but you can apply a slight undervolt if you want. That will help out slightly with lowering the core temperatures as well.
@rekareaper I thought long and hard about it and I'm really building a gaming pc. I canceled the intel i7 14th gen cpu and I'll return the motherboard, and I'm getting the ryzen 7 7800x3d. I shouldn't be spending all this money then troubleshooting something that everyone has a vague idea of what the solution is. I'm not buying a problem and then worrying about how I'm going to diagnose, fix, or return it. I'm also sure the resale value of these CPUs are going to go to shit years from now.
@@JayUchiha17 nothing wrong with that. The X3D CPUs are great for those who don’t want to tweak anything and just run them out of the box.
As long as you use a bios with the 0x129 microcode, the issue is resolved. Just don’t use the auto OC features on the boards as those compound the issue and have never been that great anyways in my opinion.
You’ll most likely be perfectly happy with either for one just gaming.
@@JayUchiha17 smart man. I was about to say you are crazy as hell if you buy intel. I’ve lost a 12900k and might be loosing my 14900k. Never again. I’m migrating workloads to MAC and I may build a new AMD system for gaming etc with the rest of my parts.
0x129 is extremely stable on my 13900k box. Z690 formula. CPU is also requesting lower voltages than before. Also, before I couldn’t run my memory at 7200 MHz stable, the system would inevitably blue screen. I had to step down to 7000 MHz. And I was having all sorts of random crashes in only certain games like Forza Horizon.
With 0x129, none of that is a problem any more. 7200 is stable. I even applied a slight overclock, a per core turbo ratio overclock. boosting to 5.9 on 1 core, 5.8 on two cores, 5.6 on 3, and 5.5 on 4 or more cores. I could probably push the system even further. But with this dialed in VIDs max out at 1.35V and average 1.24V.
For me, 0x129 is what my system needed and it is what I shall use until it’s time to sunset my system for a newer Lake. Or possibly an x3d.
The fact you are running overclocked memory controller at that frequency is impressive. Solid reliability for sure.
For me the new microcode actually made my ram unstable. I had to adjust my timings to make it work. But yes the cpu voltage is better now, before my cpu needed 1.31v for 5.6ghz all cores, now i needs 1.275v
How can I undervolt it with new microcode?
@@angelskeet1532 go to bios and set a vcore voltage. Dont forgett to lock all cores to same speed. The two boost cores dont add any performance increase in gaming or multitasking. Then you can try something like 1.35v and lower the voltage until you find the lowest possible stable voltage.
The apparent rule of thumb with the 13th gen CPUs is limit the voltage at 1.4.
not on msi boards. unfortunately I bought a z790-a board from them, and I have no option to limit voltages. needless to say, I will never buy an msi board again. not only because of that, but also because of the endless amount of bugs with their bios.
@@jacobgaysawyer337 well that's distressing. Because the new board I bought is an MSI tomahawk. Good to know I guess....
@@Stormpriest only advice I can give u is do not update the bios unless you are 100% sure the version ur about to use is clean of bugs. my experience was.. insane to say the least.
@@jacobgaysawyer337 bro wtf are u talking about. First of all, yes u can, secondly, have u ever heard of xtu? lol this is nonsensical to talk negatively about Msi mobo’s because of ur incompetence. Jesus Christ.
@@jacobgaysawyer337 u do know u can go back to old bios versions right? Or maybe ur unaware of this feature as well…
Motherboards can "request" for over 1.55v for the CPU. The Intel Microcode fix will allow the CPU to ignore unreasonable power requests.
The patch was to fix voltage spikes that you cannot see with the motherboard sensing software. You need an oscilloscope probed to the voltage rails of the CPU to see it. In many cases the voltage pre patch was going above 1.6V for a few microseconds, but it was doing it very frequently usually between a high power state to idle where the CPU was requesting the wrong voltage, this is made worse in some cases depending on the power delivery components on the motherboard.
Increasing the voltage overall might help with stability if it was unstable to begin with. Will it decrease the life of the CPU, maybe. 1.5V does sound rather high, AMD CPU's are 1.35V to 1.4v, but I assume it's entirely on the process used by each company (Intel vs TSMC) to what each can handle. The oxidisation issue was limited to the 13th gen on batches between a certain date range (that Intel have no disclosed).
The damage to the CPU is caused by a process called electromigration, you can read about that on Wikipedia or other places.
Highly appreciate your very relevant note on electromigration. For others who see this, a picture is a thousand words, and you can see the effect on the wikipedia page.
Intel users dealing with instability and micro codes that increase power draw, while AMD users downloaded 10% extra performance for free!
Edit: Guarantee that Intel chips will have more stability and degradation issues after this "fix", and will have to come out with a new fix for the previous fix. RMA nightmare.
"while AMD users downloaded 10% extra performance for free!"
Let me fix that: People freshly installing a preview-version of Win11 get the performance-increase of disabled memory security.
What do you mean? Is there a new driver update for AMD users?
@@ABaumstumpf you don't need a preview version. The fix is also part of the last 23H2 update.
I had memory integrity disabled from the start. Personally, in two different cpu limited games, I’ve seen 7-15% increase in performance with a 7600x. And you don’t need the preview version, there’s an update for everyone to download. 👍
If you watch the latest HWU you’ll see just how wrong you are about the memory integrity option.
Lets rephrase for you and for correct language. You get the 10% missing originally back, which leave you still losing by much the cpu race.
The bios change brought some changes to the bios, but board partners changed other things at the same time.
Some changed the AC and DC Load line Ohm settings at the same time and some did not, leading to a very messy situation, because the change changes the CPU voltage.
I love Intel, but 2 of my 13900ks failed. That was with undervolt of 1.28v, cores pinned to 5.5 ghz and 4.3 ghz (P-cores/E-cores), and no boost to 5.8 ghz.
Both failed at 6 months mark, at same exact way. Both were brand new sealed in the box.
I moved over to AMD, because I could no longer diagnose the issues
What happened when they failed? I have been running fine with my 13900ks for 2 years but since last month im starting to get kernal 41 error. Hard reset. Not sure if its the new bios or if my cpu have now failed to work properly.
Its passing all stress test for cpu, gpu and ram but still crashes randomly.
Try adding a little more voltage to your ram i had the same issue with my 13900k so if your xmp is 7200mhz at 1.40 volts move it up to 1.41 volts
@@betawing1914 will try thank you!
Dumped my 13900k and 13600k. I’m tired of getting no fix of the issue. Got my 7950x3D, zero issue now, and I can run with eco mode, so much more energy efficient.
@@betawing1914 yeah it was my ram, tested yesterday. Apparently the new microcode bios did not like my highspeed ram, had to adjust the timings. 🤦
For Intel, what better way is there to push users into their new processor line than to time bomb the existing ones?
One detail has stood out to me since the 0x129 microcode release before I heard anyone say anything: Intel didn't say it was a fix, they just said it "addressed" voltages. Therefore, it's not a fix but a method of keeping performance as advertised and merely slowing degradation.
well, degradation is not reversible and, at some point, inevitable. The microcode was mismanaging voltages, making the processors degrade faster than advertised. If, and that is a really big if, the new microcode keeps degradation times in specs, then it is the only fix they could ever provide.
@@Niyucuatro I look at it a little differently. It's the only fix they could provide that would keep them from being sued performance-wise before the warranty runs out which allows the cpu's to degrade slower than before the update while not actually fixing the problem of degradation. Intel has to define what the specs (voltages) are supposed to be (they haven't, so there are no specs where it really counts), keeping in mind the fact that if any microcode update slows the cpu down enough to make it be below the advertised speeds it could and probably would be considered false advertising performance-wise with impending lawsuits that Intel doesn't seem to be able to afford considering they didn't do a recall of the affected raptor lake cpu's. So to summarize that, I suspect that they aren't willing to say "This microcode is a definite fix" (for all cpu's that haven't degraded) because they can't lower the voltage enough because it would hurt performance enough to be sued. They're walking a very thin line here and I suspect that they need to lower the performance even more to make them stop degrading except they can't.
1.72V?? WTF if your cpu ever reaches that, its probably dead within days or weeks.
That's not really how CPUs work. I can upload a video running my 13700k at 1.8v if you like. Worst case scenario instead of the CPU lasting 100 years, now its only going to last 95 years. You will not notice it.
@@juice7661
Clearly Intel CPU's barely last 2 years so your point is moot.
@@SMGJohn Source?
@@juice7661
Are you asleep or what? Or maybe you are in denial.
@@SMGJohn I see no source🤣
Since the moment i got my 13600k, i noticed high temps. Even with a nice water cooler i was getting around 100ºC. That was unacceptable for me. I reduced voltage and performance core ratio to like -0.150v and 47x, directly from Intel xtreme tunning software. I lost performance on my benchmarks, but now my cpu gets 78ºC max. For a long life cpu, that works, but it's too bad to know that my cpu is limited. I did this a year ago btw.
I haven't seen my 13600K go above 60 degrees during heavy normal use like cpu demanding games. It goes up to 80 degrees briefly during Intel Tuning benchmarking. Yours was probably defective out of the box.
Should we purchase 12 gen cpu?
They can't fix it. The only fix is a complete recall which would render Intel bankrupt.
Bro it was 2+ weeks from the day I started my RMA for a new 14900k. The cross shipping form I sent back, Intel tried to tell me there was an issue with my info and management would get back to me. 2 different promised call backs that never happened later, and never being able to since reach anyone on the phone with my ticket number, even after waiting over an hour on hold multiple times…via the Intel chat on their website, I threatened to and showed them the documentation for 1 of the class action law suits I was going to sign up for, I then received a phone call within 15 minutes and my replacement cpu was shipped to me with next day air UPS shipping….lol this is a true story.
Did they make you pay for the replacement?
@@_Mattz1nho I had to give them my credit card and they charged $450 and held it hostage until they received my old cpu. Although, because I complained so much about how much they suck and how long they ignored me, they ended up covering the $25 shipping charge lol
the worst part is that it is intel, so it must be true and others has similar or worse experiences, terrible company.
@@_Mattz1nho they charged my credit card for the price of the cpu until they received the defective one. Trust me, u WILL need to follow up about getting ur money back or else it will take up to 30 days.
Best to use modified settings with 0x129 BIOS. I did analysis on Asus ProArt Z790 with 14700K: ua-cam.com/video/J2k9_dqKlnw/v-deo.htmlsi=f8MFkxBFxE7U625c at 17:00 in video is where I show comparisons. Bottom line is (and probably for other CPUs as well) with a few settings you lower temp/power/vcore substantially, with almost no performance hit, and there is no thermal throttling. With 0x129 BIOS as is there is thermal throttling, vcore still goes up to 1.5 something. Modified settings keep things more under control which is just another means to increase likelihood of stability/longevity.
Had a DDR5 4 DIMM configuration (yes I know its not recommended). Was working fine even with XMP at 6000 Mhz. I then updated my BIOS with the new 0x129 microcode, got instability, bluescreen, crashes, FPS drops etc. I removed 2 RAM sticks and disabled XMP. Changed to "Performance" profile in BIOS and it is stable again with some rare crashes etc. Sooo frustrated, so I ordered an AMD CPU now.
Hello, if I just bought a new PC that has a 1400KF, should I first test it to see if it has the overheating issue before updating the BIOS on my motherboard? or should I just update it right away?
Appreciate the effort you took into making this video but it is extremely badly researched and your voltage numbers are not correctly measured, nor are they relevant to what the actual problem is. The voltage spikes seen in the older microcode were in excess of 1.6v, and they were not seen in the middle of heavy, multithreaded workloads, they were seen during transients of applying/releasing loads. This is in line with some of intel's communications on the matter.
As others mentioned, this fix will not help CPUs that have already been damaged, it will only stop (further) damage to CPUs that are still functional within the operating margins.
I was afraid this was a problem, I have a 13900k and I bought a used RTX4080 about 6 months ago. I thought it was the GPU. The system just randomly boots. The windows event viewer shows nothing. I did the microcode update and it's still happening, at this point I don't know what to do?
7950x3d
@@dubya85 Sounds good to me. Are you going to buy it for me?
@@derodge 😆👍🏼
15:30
i eliminated the issue and fixed all crashing and stability issues....
i switched to AMD.
It's funny how there are double standars. Some reviewers don't use PBO on AMD, saying that it's not the default profile, and therefore limit its performance. But when it comes to Intel, they are willing to try out 10 different profiles, under volting and using massive water coolers, while saying that's normal.
did the same thing
@@abaj006 I had to do so much settings In the bios in an attempt to make the intel chip stable (and it still wouldn't be) or become less hot, that I'm used to it.
When I switched to AMD I heard about PBO, which made the chip run cooler, with more performance. I was sold immediately. Easiest "overclock?" Ever. And I always changed these settings conservatively. I'm nit pushing the limit, all I want is for it to run a little bit cooler, so I don't have to replace it as soon.
Love that remedey
Amd has come along way in both gpu and cpu. Biggest thing was getting there driver thing worked ouy
In this video I just saw a lot of misconceptions on how modern CPU operates: 1. what you refer to as high temperatures is absolutely within spec; 2. the UPS beeping doesn't have anything to do with CPU issues, if anything your UPS is underspecced; 3. Vcore is not what the CPU gets but also what the CPU requests, there are multiple independent power lines going to the chip; 4. your high voltages are also within spec as stated by Intel, and while it's not normal to have them in idle it's absolutely expected to reach them under load; 5. when you say the CPU is throttling it's actually not as it's still over the baseline clock, it just limited the turbo-boost and this is absolutely normal.
If there was a problem with high voltages under certain conditions that got corrected with a microcode update, it doesn't mean this problem was present under all circumstances so it's a wrong expectation to see your voltages decrease, you might have never experienced the high voltage issue that Intel fixed.
i brought my 15-13600k in january 2023 - havnt updated the bios - do i update the bio even if i have had 0 instability issue - it does run hot yes but no crashes at all
I havent updated, but ive never had a problem with my gen 13 cpu. Were there more specific batches that had the problems or the entire run?
I recently updated to the new microcode. Im running a 13900KF on a Strix Z690-G. Intel Extreme Profile, Enforced Power Limits, MCE off and Intel Fail Safe VID behavior gets me 35K on cinebench R23 and 1.50v max VID. Otherhand, Asus Advanced Profile, Disabled Power Limits, MCE on and Auto VID behavior gets me 40K in Cinebench R23 and 1.42v max VID
hey i got a some question please answer
1. does the cpu affect laptop cpu?
2. what are the models of processer that are affected by the cpu issue?
laptop is fine, only 13700k 13900k 14700k and 14900k ks model affect.
I bought a 13600 k last week, its installed on z790 asrock pg lightning and i have aio id-cooling 240. I have no issues with default bios, temps are on idle 26-28 degrees, and on gaming 40-50 degrees(summer) I cant complain for now.
the OS also updates microcode via updates, those are always applying the newest, coming from the bios or the OS, you should check always the current microcode of the running CPU while testing versions
Cool advice : If you just come to get a brand new Core i9, undervolt it while it's not too late...
Hi @TechLens - can You write if betwen all that problems in cpu stability did You notice any problems with booting system? For example: explorer.exe is nit automaticly starting after log in to system or even worst the logonui is not showing after initial system boot? Any common error logs that are repeating when cpu crash?
Many years ago, when we were in the 32nm range, 1.5v was the voltage where CPUs started eating themselves. I can't imagine that shrinking geometries has raised safe voltages. I would rather operate in the 1.3v area.
A lot of the crashing was due to motherboards undervolting due to load line calibration issues. Where the microcode issue was causing transient spiked above 1.55v which appears to be the danger zone. These are two separate issues. Undervolting average voltages and overvolting in transient spikes. Hence the intel fix is supposed to both raise average voltages to improve stability and reduce the peaks of the transient spikes so they don't go above 1.55v. Addressing both the overvolting and undervolting. This appears what you are observing.
Raising voltage and reducing power limit is of course going to impact performance. But you can manually tweak this with an undervolt and recover the performance.
TechLens : the safe continuous voltage for the 12-th, 13-th and 14-th GEN Intel cpu-s is 1,30-1,35 MAX. All voltages above that are dangerous the moment that they become more than just nanosecond peak ones. The problem arises when higher values of current are pumped into the cpu and in correlation with the 1,30+ Voltages the CPU starts to draw abnormal amounts of PWR. The damage occurs at that point but it also occurs even if little current is pumped into the cpu but the voltages are continuously higher than 1,30 V.The degradation is a result of certain portions/clusters of tranzistors literally being melted together.
thank you for this video and sharing your experience, just one question, if VIDS voltages are what the CPU are requesting why we focus so much on monitor them instead of monitor VCore which is what was actually delivering? what I am missing here
I'm pretty sure intel is just trying to cover it up until the CPUs are out of warranty range because a recall would make them look REALLY bad. That would explain why the voltages are boosted. The performance is already degraded so if all they have to do is boost the voltages then they'll do that knowing most people will blindly apply the firmware updates. This is a large corporation which is suffering pretty big losses and has a history of shady business dealings. I don't know what's actually going on behind the scenes or anything just speculating with what's publicly available.
Yeah, that’s why they extended the warranty by 2 years. Lol.
With software-readouts you can NOT check for voltage-spikes. So just from that alone your "testing" is worthless.
And the 0x129 is just there to enforce the 1.55V VID-request limit. The UEFI actually has a lot more protection-features but most motherboard-vendors had decided to disable them all.
1.55V is still far too high. Anyone experienced in overclocking will tell you that is a very dangerous voltage that WILL brick your chip eventually unless you have the goldenest of golden sample silicon.
@@asdf_asdf948 "Anyone experienced in overclocking will tell you that"
Your assertions are wrong.
1.55V is really high, however the thing that will cause damage and why high voltages are avoided are the spikes that go beyond that. If you lock your CPU to say 5GHz allcore at 1.55 V (As people did and still do for that) then you will not have problems aside from the CPU constantly wasting power.
And the high voltages only get applied when you get 1-2 cores to full load - and then the problem is still when the load actually ends and you get overshoot.
Just one question as owner of an I9-13900k myself who has experienced random game crashes and even reboots in the past myself though it always was rathre random. Did you try to assign a limited set of CPU cores, like only the P-cores to a game and remove the other cores, like the E-cores, from the game and let it do other stuff instead? Did that help reduce the power consumption? From what I noticed, this seems to make games more stable like in Star Citizen, which used to be very CPU heavy before their graphic updates, Red Dead Redemption 2 or Hogwarts Legacy.
I'm having a tough time optimizing my I-914-900K. I just got a replacement from Intel as I have had nothing but issues using the ASUS motherboard that automatically overclocks dark hero x5790. Even turning that off completely doesn't stop my CPU from thermal throttling. It never overheats but it thermal throttles every benchmark that is CPU intensive. Not sure what I should do if I should undervolt it or leave it at stock or overclock it
I had to RMA my 13900k to Intel I was upgraded to a 14900k due to supply. I shipped via UPS and had my CPU back in 7 days . I bought a contact frame and Kryonaut Extreme and haven't had any issues after bios update.
Hi, do you know of any work-arounds for laptop users who have an intel 13th and 14th gen CPUs? Let's say if their laptop doesn't have a problem and wants to at least avoid the problem that others may have on Intel's ongoing issue.
I have recently bought 13900HX Laptop. Its been a month now that I am using it. Played many games since then especially Black Myth Wukong. I don't see any crashes on it. Hopefully laptops are not really affected.
@@blenderesque708 did you do anything with it, such as update the laptop's BIOS, download Intel's microcode, update the laptop's driver from the manufacturer's website?
Class Action Lawsuit For 14th & 13th Gen CPU Instability Issues Begin. I hope we all get the money back from these scammers!!!
if you need more voltage to fix crash, it's not a good sign, how long will it last that cpu? and how much power draw required for that shader compilation? 400watt?.....can you do a video with undervolting? because doesn't seems a fix but just more voltage to the ones who has a degraded cpu
Will newer CPUs fabricated after this was detected be safe to buy? How to tell?
i thought the update was to fix extreme voltage spikes with TVB not general voltage to the cpu?
yes that's right, it did also add a 1.55v VID limit when using the Intel pre-sets (that should be the MB boot default now). Also set 253W for the PL1 and PL2 400A for ICCMax (it actually did this part in the least update too) previously they were often set at unlimited!
how do i get that update?
Idk how this happens.. ive had 13900k since launch and have done the bios update for micro code fix. No issues at all. (Pre and Post Update). Temps dont sky rocket either. Or voltage spikes. (I dont overclock)
If they do Individual Calibration, why not tell us the result, on the box with each CPU?
So many utubbers teaching how to build a system incorrectly doesn't help at all either!
I recently updated my bios to the new microcode, for some reason MSI took there sweet time releasing it for my motherboard. The big thing I noticed was the clock speed was way down. I have a i5 13600k which is supposed to be 5.1GHz. Now the best it can do is 3.5GHz
Set your settings to factory defults after bios update and re apply all your settings.
@@fofal Did that and still acts off. I'm not getting the stutters I was getting before. Doing work related things, or having multiple windows open I was getting a lot of stutters and lag. But now the speed and voltages are all over the place. Default settings don't change it. Undervolting I can get it to stay closer to 4.4GHz, and using Intel Extreme Tuning's automatic overclock I can get it to 5.1GHz. But again the speed jumps all over the place. Maybe the cpu is defective and it's in it's early stages, I have no idea. Never had an Intel cpu act like that and i've been using them since 8th gen
@@spentcasing3990 RMA it you might get a 14th gen.
@@fofal 14th gen is just as fucked as 13th gen. Maybe more so because of the higher voltages. I'm saving up and switching to AMD
@@spentcasing3990Is it overheating? What is your long duration boost (PL1)? I assume previous BIOS' undevolt caused your chip to run at its edge of stability hence the stutters
When you take a deeper drive under the hood...
Exactly how are you taking this deeper dive under the hood and does it involve something more than software that monitors voltages and cinebench?
Just bought this CPU refurbished on eBay. It arrives next week. Should I return it? I'm scared. Wish I saw your video earlier, any advice?
yes return it and get a 12th gen processor if you want to stick with Intel. It's the latest generation that doesn't suffer from these problems. Why do you even think the guy on eBay was seeling it at this point in time? Also wtf does refurbished CPU even mean? You can't refurbish a CPU as far as I know. If you can't go 12th gen Intel (which btw has the same LGA 1700 architecture as 13th and 14th gens so it fits on the motherboard for those gens as well), go AMD
@@George-Chris the whole Computer is refurbished not the CPU. I was hoping the update would fix it. Damn this is gonna be a pain returning it
Not every chip has the problem.
Recently replaced my CPU and mobo, went from Intel to AMD, and no regrets. What I would also caution against is using Intel's RAID, as that made the switch more difficult than I want in the future. Just an FYI.
Im hesitant to upgrade. My two month old new pc runs great and plus it’s an i5 14400F and temps and volts never get out of line. I haven’t heard if the i5’s are affected
Though I have experienced no issues (on ASRock Z690 BIOS 13.05, now two years old), my 13700K remains in stasis while reviewers get this fully sorted out. I know that at some that point I'll have to fire it back up, apply the latest BIOS/microcode, and see what happens.
To sum it up, everyone wants golden sample running at lowest voltage.
i had no prob with gigabytes new bios microcode update. and perhaps you look at wrong type of voltage elevation. it requests specific voltages at specific times
SO IF I HAVE NEW 14600K WHAT AM I SUPOSE TO DO DUDE SHOULD I UPDATE BIOS OR NOT ?
TLDNR YES - update to Bios with 0x129 microcode.
Then if you are comfortable with changing Bios configuration select the Performance intel setting if you want to be a little more conservative (it basically lowers your power form 400A to 307A) or stick with the Extreme setting (default) if you want best performance. It might run a little hot but changing anything else to optimise it needs some knowledge, check out some other streamers too on the subject. (hint - It will involve AC/DC Load line changes and CPU negative voltage offset ( undervolting))
I’ve noticed a few things with the new bios for my i9 13900k on my strix z790-A. My temps are higher, games that were butter smooth now crash and have stuttering. My temps were very good under 45C during heavy gaming. Now playing the same games I’m hitting 60C+. Games that wouldn’t even launch before (vram error) now launch properly but have Intermittent crashing to desktop. I spent WAY too much money for this to be happening
It's current that causes physical damage. You gain stability by increasing voltage and decreasing current (among other things). To decrease 'average' current you can decrease frequency. Because of the reduction in frequency, less work is done per time; you must run benchmarks to measure work rate.
I have always been on the safe side, all the time my multi core option on asus was turned off.
Lately I have been observing strange things, performance drop after undervolt, different results, on the OCCT application at 100% load I do not reach speeds above 4.8 MHz Processor + Ram 5.2, on geekbenchmark it is solid 5.5, with turbo 5.8
We are talking about 13900K
On cinebench R23 on multicore I cannot reach 37K mostly around 36600, 36800
A strange smell also appeared a week ago, it was for a few minutes but it's strange
It's hard to say even after tests what condition the processor is in, and the motherboard is also unknown in what condition it is after such loads.
Well the intel info said "Address high voltage" now "Lower high voltage". So maybe they did some tests and decided they'd want to keep the voltages up so they wouldn't tank so much in benchmarks.
I'm not a Pro in this, just a newbie. Fearing supply chain wars from China, and having an aging but working original 14 year old i7 860 HP Pavilion computer, I built myself a gaming PC with MSI's Carbon WiFi Z790, i9 13900K, 4090 GPU (both liquid cooled) and GSkill DDR5 5600 32GB x 4 (128GB) RAM. Before I updated the microcode my XMP profile showed 5400MHz RAM speed but 5.3GHz CPU speed with the CPU in "Game Boost" mode. Now, after the x129 update, the CPU now runs at 5.9GHz (and the RAM runs at the same speed, i.e. 5400MHz). So the question is: which is better for stability and performance: higher or lower CPU clock speed? I checked my voltages compared to my previous setup (from last year's BIOS), but they haven't changed much (maybe 1 or 2mV). I haven't benchmarked my computer yet.
The i9-13900k is 50% off here and it is still too expensive considering the risk. I have no confidence in Intel CPU’s anymore.
Can u get one used
The performance on the latest i7 and it’s was so impressive that it’s such a shame this whole malarkey has happened. Mistakes happen, but to push it out to the public when you know there are such issues, and then when people RMA the product for valid issues just to shirk responsibility, and then not to fix the issues from happening, and then to not be transparent at all about the issues is simply malice, rather than incompetence. It’s so depressing to see these great American companies like Boeing and Intel decide to profiteer instead of innovating and ruining these companies chasing short term profits.
It's a red flag to read voltage upwards of 1.45v but 1.72v? I don't think I'd ever buy another intel chip again, I wouldn't care how much they tell us that this is normal behavior. 1.72v.... unreal!!
only fix that worked for me was disableing turbo boost, turbo boost 3.0 in the bios and windows 11 turbo boost!
Any *tutorials on reducing the voltage* for this particular situation?
Yes, but you didn't say for which specific cpu because you don't just lower voltage, you have to lock core frequency too to keep it from boosting too high at stock frequencies requiring more voltage just to not crash.
@@pf100andahalf 14900K, please.
In his case I assume it was some crazy AC LL shipped with the new bios, like 1,7 mOhm (it happened to me after updating my biose I got 1,4V instead of previous 1,3V before any undervolt). Typical AC LL should be between 0,01 - 1 mOhm depending on your LLC (the steeper the curve, the more AC LL you need to compensate, but it should never go past 1,1 mOhm).
Seeing that you not only use a UPS with your PC, but you have it affixed to your desk /respect.
I've been using them for about twelve years now. Initially out of necessity due to living in an area that had constant brownouts, but then just for peace of mind. My current residence doesn't have constant brownouts, but it does have power fluctuations every few months and at least two times a year the power goes out completely for about eight hours.
This has been on the stove for a few years now. Ever since the rise of Ryzen, Intel has pushed crazy power into an architecture that would eventually fail. Which it has. Their “P” core / “E” core experiment has been a huge failure. There was no possibility of keeping up to AMD’s core count on that architecture and I’m sure there were some Intel engineers who voiced their concerns………but were ignored. I’m not sure I will build Intel rigs for a while moving forward. There is no trust remaining.
3:46 - most likely just combined power consumption of your cpu + gpu with unlimited fps. For me, even shader compilation takes 20% less power than CB R 23 and 35% less power than AVX2 benchmark like OCCT Extreme.
I solved the problem by capping the clock of my 14900K to 5.4 Ghz. There is no other way around it. Wattage = Voltage * Amperage. You can't have 6Ghz on this architecture, period.
Waste of money really. Really bad a look for intel. A cpu that can do 5.7 all core and more in some cases has to be capped at 5.4 just to work. RMA it brother
There are other ways round this. 0x129 has stopped the damaging/ crashing spikes (1.6v+) but still set voltage at 1.55V which is too hight in my opinion and could still degrade the CPU over time, but has meant less crashing.
So alternative to locked frequency is you can go into the bios and set a max voltage yourself, (say 1.4v) then under volt the CPU, -0.100mv offset in my case, your CPU might be able to take less or more, trial and error. That gives you full range of frequency’s (inc 6000mhz boost) and the system never goes over 1.4v. (oh and thermals are better too)
@@Harrygostyou can't do both lol. fix voltage or minus offset
@@SomeoneElse6282 Mine does work at stock frequency. This is just me being extra careful
@@nivea878 Not literally fix voltage of course, bad terminology on my part sorry. set voltage max with 1.4v VID limit (IA VR Voltage Limit)
Just check your AC load line. MSI for example is pumping 1,7 mOhm for some boards with the new bios. It means +170 mV per 100 A minus LLC impendance. Max LLC impendance is 1,1 mOhm for most boards. If you are using LLC auto (steepest curve) your AC LL should be within 0,8 - 1,1 range.
I don't mean this to at all downplay this issue, I am asking in sincerity.
Do some chips just luck out and not experience these issues?
No, all chips age with use, its just that lower SKU Bins running with bad bios settings will show problems (degrading) first. It seems to me that what Intel are trying to do is get rid of the bad bios setting problem and a microcode update to stabilise the problem so your chip lasts at least 5yrs (warranty) If you a home user you should get 10yrs out of your CPU, unless your running a server farm or mining 😁 Hey you don't run your Car with the foot to the floor do you 😜
@@Harrygost 👍🏼
The retailers in system integrators are doing the right thing by extending the warranties. They are going to be hurt the worst. They won't recoup the time and labor, shipping cost to replace all these over the years. The margins and profits can't balance against this future level of repair.
Is there any fix to prevent overclocking and crashing with certain software and games, using a computer with an Intel chipset? I haven't noticed one yet that's actually legit, so I'm almost at the point of giving up and buying a new mini PC with AMD (If I can find a good one that is).
Meanwhile, my 12 years old Core i7 3770K runs @ 4,5 GHz with a Vcore of only 1,254 Volts. And I can still play Cyperpunk 2077 @ 1080p max settings @ 85~88 fps 😂
Ivy Bridge FTW, mates 😁🥰😂😂
Very good architecture
When I got my 5600 it would run at 1.5V+ default on my b350 f-strix and wouldn't hit 4.6 GHz singlecore pbo only, reaching ~4600 in cpuz.
After me overclocking, I'm now running 4650mhz all core @ 1.2 V + LLC goes up to 1.263 V hitting around 5100 in cpuz.
Still lower temps on the vrm and CPU.
I feel like they are all trying to fry our hardware, so we can buy new...
bios 0x12B solve all... Got more perf, stability, better temp... install asus suite, launch Ai OC, keep scale at 100, and when you reboot, ai suite will update correctly your bios. After you can disable it. I have disable TVB for max Perf. Got stability, perf (6ghz) and good temp.
The MAIN problem is that intel is calling this a FIX but this is not a fix. If your bread is already cook it wont uncooked it. This will only prevent your current uncooked bread to not burn as fast.
They use "fix" as a word because if they didn't, no one would download it period, even if it does fix the nasty voltage spikes above 1.55V like how it does here. If your CPU was already kicking the bucket, this won't fix it. Nothing can fix an already degraded, about to die CPU ya know.
I'm happy I jumped from 8th gen 8086k to 7900x3d
Easy way to not hit the high vid request on the new baseline and microcode update is say lock your two preferred cores to the rest like a 14900k is 57, then up your IA VR to 1400-1420, that should give you just enough voltage to have all your p cores to hit 5.7ghz since they will have the power to do so your VID will top out at 1.36X-1.38Xv, then the Vcore will also stay around 1.35V. It could take less depends on the quality of your chip. My P and E cores are not that great, so need a little more for my chip specifically. Do not want to do all of that you can just disable your E cores just expect your load times to take a little longer.
I want to build a new PC, but am totally confused, don't know what to do. The GPU (NVIDIA) problems with the connectors, and Intel with the problems. I have had the Intel CPU and NVIDIA GPUs since the early 2000s. I want to keep it that way, I know nothing at all about AMD, and don't really want to, I believe :)
Nvidia connectors are fine, i've had a 4090 from the day it released and all i did was make sure the connector was fully clicked into place.. also since about 9 months ago Nvidia made the connector a little shorter so they have less wriggle room for people to not plug in correctly.
@@_Mattz1nho Anecdotal evidence is hardly assuring that a statistical problem doesn't exist. Even if you don't know anything about how many of these have issues, just looking at the new connector compared to the previous connector, it's remarkably undersized for the increased power draw.
My Coffee Lake CPU doesn't suffer from any of these issues. 700mhz all core overclock constant at just 1v to the CPU, under load my CPU barley pulls 30 watts CPU Package Power as a whole. It still bangs out 90 fps at 1440p gaming in new titles. Sure it's old now but it works so why replace?
Don't get e wrong i am in the market for a newer CPU MOBO combo, but not 13th or 14th gen intel and i do not like the fact that X3D chips do not like to be pushed beyond stock which is something i have always enjoyed getting more from. As the years passed the stock 4ghz was not enough so here we are with a good overclock and it cost me nothing to extend the rigs life an extra 3 years or so.
There is no need for a new CPU no more. So they fix it by destroying yours purchase in a slow burning process
i9 13900K
-0.03 offset
PL1 125w
PL2 225w
LLC mode 4
Sorted.
Software is NOT an issue. I tried windows 10/11/Windows Server 2022+2025Preview & (5+ versions of Linux with different base) of which most crashed or froze during install. I tried new SSD and different RAM. I tried with NO GPU and with old 1070 GPU. At the moment I have to run CPU at x50 multiplier for it to be stable. All of the above I tried while rotating hardware around and yes DDU was done in safe mode. My place of purchase refuses to do RMA because it passed their 24h street test. My crashes are with normal work, office stuff which is about 95% of what I use the 14900k for. I Work in IT and can do all this stuff while sleeping. That's how easy it all is. BIOS is up to date
When you say "the values are something they check" you assume that the data is ever exposed to humans. There is no reason to believe that anyone at Intel has been looking at the values or their range. If Intel failed to monitor the values applied by automation and generated by formula, the voltages could have crept over time. Without a defined upper maximum, unlimited drift could have occurred over the years. Process changes over generations drift beyond the initial design spec for the code which optimizes and one day you discover that a blight has been creeping into the product for years. One of the dangers of extreme organizational complexity is that knowledge becomes compartmentalized and quickly lost. The old process may look like it works for the new product but you can only tune for issues which you either foresee or detect. If you hid a variable in the process which drifted over time , you are blind to it.
VID table from the factory has nothing to do with max safe voltage.
I have used AMD for the longest time and 5 months ago I built a computer that my company spent 5 grand on and I decided to go with Intel because of supposedly raving reviews. What did I get in return? constant crashes, weirdly slow computer almost all the time, even very light programs (like vcarve or illustrator). This is ridiculous. I should have went with Ryzen series or a threadripper. I really screwed my self here.
Intel is still working on microcode fixes. RMA your 13th or 14th gen processor for a full refund and buy a 12th gen. the 0x129 microcode doesn't fix the root cause. Perhaps the next microcode version will fix the root cause, but at this point just request a refund.
Intel's only hope is Core Ultra. But everybody is going to call it i9-285K and i7-265K anyways
I'm very interested in seeing how the Bartlett Lake CPUs turn out. (I'm riding LGA1700 _until the wheels fall off!_ 🚙😜)
With these series of chips, we just need a do-over, this is such a mess. Intel is just stalling to avoid being sued. As long as they can say that they are doing something, they stay out of court. It seems it isn’t a matter of will these particular chips die, it’s when and will you be the in lucky person that will be left holding the bag if it’s after your warranty.
Monitoring voltage is great advice after finding out how different each individual chip is. It is truly great that the Information Age has provided media outlets the ability to uncover what they use to hide. Great job here and outstanding coverage from a different prospective.
You seem to be assuming only one scenario is possible - either more voltage is being supplied or less. But I think what's actually happening is that the CPU is getting both more AND less voltage, as needed. What do I mean by that? I think a cap has been placed on the maximum amount of voltage the CPU can request, because super-high spikes were causing degradation, AND in parallel these CPUs are now being fed more overall voltage (minus the big spikes) to ensure stability. And by doing that, Intel is theoretically preventing degradation while also increasing overall system stability. Will it work long-term? Only time will tell. But for now at least, Intel believes 1.55v and under is safe - safe enough to get CPUs past their newly extended warranty. Personally, on my own 14900k, just to be extra safe, I've installed the new bios with the new microcode AND I've capped all of my P cores at 57x and applied an undervolt.
I am an Intel fanboi.
That being said, any firmware or software-based solution, will not "fix" a problematic manufacturing flaw. At best, firmware may moderate performance issues in order to avoid using them. As for the overvolts, well to allow for performance that is at parity with the expected norms, it seems logical that decreasing efficiency of some paths would require some level of compensation by increasing efficiency of others (I would expect).
To be clear, I am only relying on logic, I have no schematics of functionality nor an understanding of what is being performed by these patches. So, grain of salt. 30 years in the industry is all I have to work with. Without a team of engineers, it's hard to speculate.
If I had to guess, no firmware patch could "fix" a manufacturing flaw. the best we could expect is to compensate for these flaws.
Interesting perspective though. thanks.
wait... did you just let your CPU fluctuate it's clock speeds and voltages based on temperature??? oh...
RMA your 14900k if you can't undervolt it anymore without crashing. Once you have your replacement cpu, just sell it, switch to amd or just downgrade if you really need intel quicksync/performance features.