Intel Ships 0x129 Microcode Update for 13th and 14th Generation Processors with Stability Issues

[AleksandarK]

Intel has officially started shipping the "0x129" microcode update for its 13th and 14th generation "Raptor Lake" and "Raptor Lake Refresh" processors. This critical update is currently being pushed to all OEM/ODM partners to address the stability issues that Intel's processors have been facing. According to Intel, this microcode update fixes "incorrect voltage requests to the processor that are causing elevated operating voltage." Intel's analysis shows that the root cause of stability problems is caused by too high voltage during operation of the processor. These increases to voltage cause degradation that increases the minimum voltage required for stable operation. Intel calls this "Vmin"—it's a theoretical construct, not an actual voltage, think "speed for an airplane required to fly". The latest 0x129 microcode patch will limit the processor's voltage to no higher than 1.55 V, which should avoid further degradation. Overclocking is still supported, enthusiasts will have to disable the eTVB setting in their BIOS to push the processor beyond the 1.55 V threshold. The company's internal testing shows that the new default settings with limited voltages with standard run-to-run variations show minimal performance impact, with only a single game (Hitman 3: Dartmoor) showing degradation. For a full statement from Intel, see the quote below.

Intel is currently distributing to its OEM/ODM partners a new microcode patch (0x129) for its Intel Core 13th/14th Gen desktop processors which will address incorrect voltage requests to the processor that are causing elevated operating voltage.

For all Intel Core 13th/14th Gen desktop processor users: This patch is being distributed via BIOS update and will not be available through operating system updates. Intel is working with its partners to ensure timely validation and rollout of the BIOS update for systems currently in service.

Instability Analysis Update - Microcode Background and Performance Implications

In addition to extended warranty coverage, Intel has released three mitigations related to the instability issue - commonly experienced as consistent application crashes and repeated hangs - to help stabilize customer systems with Intel Core 13th and 14th gen desktop processors:
1. Intel default settings to avoid elevated power delivery impact to the processor (May 2024)
2. Microcode 0x125 to fix the eTVB issue in i9 processors (June 2024)
3. Microcode 0x129 to address elevated voltages (August 2024)

Intel's current analysis finds there is a significant increase to the minimum operating voltage (Vmin) across multiple cores on affected processors due to elevated voltages. Elevated voltage events can accumulate over time and contribute to the increase in Vmin for the processor.

The latest microcode update (0x129) will limit voltage requests above 1.55V as a preventative mitigation for processors not experiencing instability symptoms. This latest microcode update will primarily improve operating conditions for K/KF/KS processors. Intel is also confirming, based on extensive validation, all future products will not be affected by this issue.

Intel is continuing to investigate mitigations for scenarios that can result in Vmin shift on potentially impacted Intel Core 13th and 14th Gen desktop processors. Intel will provide updates by end of August.

Intel's internal testing - utilizing Intel Default Settings - indicates performance impact is within run-to-run variation (eg. 3DMark: Timespy, WebXPRT 4, Cinebench R24, Blender 4.2.0) with a few sub-tests showing moderate impacts (WebXPRT Online Homework; PugetBench GPU Effects Score). For gaming workloads tested, performance has also been within run-to-run variation (eg. Cyberpunk 2077, Shadow of the Tomb Raider, Total War: Warhammer III - Mirrors of Madness) with one exception showing slightly more impact (Hitman 3: Dartmoor). However, system performance is dependent on configuration and several other factors.

For unlocked Intel Core 13th and 14th Gen desktop processors, this latest microcode update (0x129) will not prevent users from overclocking if they so choose. Users can disable the eTVB setting in their BIOS if they wish to push above the 1.55V threshold. As always, Intel recommends users proceed with caution when overclocking their desktop processors, as overclocking may void their warranty and/or affect system health. As a general best practice, Intel recommends customers with Intel Core 13th and 14th Gen desktop processors utilize the Intel Default Settings.

In light of the recently announced extended warranty program, Intel is reaffirming its confidence in its products and is committed to making sure all customers who have or are currently experiencing instability symptoms on their 13th and/or 14th Gen desktop processors are supported in the exchange process. Users experiencing consistent instability symptoms should reach out to their system manufacturer (OEM/System Integrator purchase), Intel Customer Support (boxed processor), or place of purchase (tray processor) further assistance.

View at TechPowerUp Main Site | Source

 

[Dr. Dro]

For Hitman 3 to lose performance but not Cinebench (and saying this from experience), that must be one heck of an oddity

 

[ncrs]

What about performance degradation on not-"Intel Default Settings" which was the shipping configuration for most, if not all, consumer motherboards until recently?

 

[Jism]

For Hitman 3 to lose performance but not Cinebench (and saying this from experience), that must be one heck of an oddity

Single core, max boost or max voltage.

Ah well - i'm not sure if this ever was a bug or not, or silicon being pushed to the extreme just to be on top of the charts.

 

[Dr. Dro]

Single core, max boost or max voltage.

Ah well - i'm not sure if this ever was a bug or not, or silicon being pushed to the extreme just to be on top of the charts.

Nope, none of it changed. Check the ASUS announcement thread, post a bunch of stuff regarding it there

 

[Wirko]

What if the motherboard supplies a higher voltage than the CPU requests?

 

[W1zzard]

What if the motherboard supplies a higher voltage than the CPU requests?

Then this mechanism will not work. i.e. if you set voltage offset +0.1 V, it should still go above 1.55 V

 

[iameatingjam]

Is it just me or could the title be misinterpreted to sound like 129 is going to fix instability issues on cpus already experiencing them, rather than just preventing it on those that aren't ( as intel has said) ?

Also noticed this line

"Intel's analysis shows that the root cause of stability problems is caused by too high voltage during operation of the processor."

Did they say that? I remember them saying this in the July press statement:

"Intel is delivering a microcode patch which addresses the root cause of exposure to elevated voltages."

Which is not quite the same thing. But I could be mistaken, maybe they said that at a different time.

 

[Evrsr]

Highly doubt this will do anything useful. 1.55v is still a heck of a lot for this process node.

Anyone running these would be best advised to limit voltage requests in the BIOS, to at most 1.4v. Even that I wouldn't feel safe with, for CPUs running many hours per day.

 

[Dr. Dro]

Highly doubt this will do anything useful. 1.55v is still a heck of a lot for this process node.

Anyone running these would be best advised to limit voltage requests in the BIOS, to at most 1.4v. Even that I wouldn't feel safe with, for CPUs running many hours per day.

The problem with that is that 1.4 V is an incredibly conservative voltage and it is not enough to hit advertised clocks on at least 98% of samples, regardless of SKU type. Let's go with the 13900KS which is what I have most experience with, this is obviously a preselected, relatively high-quality i9-13900K.

The last and fastest of their kind – Intel 13th Gen Raptor Lake Binning | Part 3: i9-13900KS | igor´sLAB

Today is finally the day – loyal readers won’t be surprised by this –, for Intel’s last and fastest LGA1700 CPU we also have binning data for you once more. Thanks to MIFCOM, we were able to run…

www.igorslab.de

My personal i9-13900KS's 6 GHz V/F point is 1.463 according to my Apex Encore. and this is apparently a resoundingly average chip amongst its kind - at SP 108, 118 P and 89 E, it's smack in the middle of all bell curves from the Igor's Lab article and meets the exact median and average results of all of their samples, according to the same, for a 13th Gen i9 CPU to be considered a KS quality, its 6000 MHz point must be <= 1.49 V, so while it isn't the best, it's also far from the worst and the textbook "it is expected that it should behave exactly like this" chip.

The 13900K/KF, 14900K/KF and 14900KS processors have a different SP scale, but their V/F curve is not that far off -the i9-14900KS in particular has a 6200 MHz V/F point all the other SKUs do not - and it's common for those to be in the ~1.525 range. Now let's think outliers, such as stinker 13900K samples... yep

 

[W1zzard]

Is it just me or could the title be misinterpreted to sound like 129 is going to fix instability issues on cpus already experiencing them, rather than just preventing it on those that aren't ( as intel has said) ?

Also noticed this line

"Intel's analysis shows that the root cause of stability problems is caused by too high voltage during operation of the processor."

Did they say that? I remember them saying this in the July press statement:

"Intel is delivering a microcode patch which addresses the root cause of exposure to elevated voltages."

Which is not quite the same thing. But I could be mistaken, maybe they said that at a different time.

"Intel's current analysis finds there is a significant increase to the minimum operating voltage (Vmin) across multiple cores on affected processors due to elevated voltages. Elevated voltage events can accumulate over time and contribute to the increase in Vmin for the processor."

This is just confusing wording to say "high voltage causes something to happen, which, over time, makes the processor unstable at default voltage, so it would only work if we gave it higher voltage"

 

[dont whant to set it"']

Congratulations,you broke it. Too late?
What did Intel expect from turbo boosting the thermal velocity boost the turbo boost in the first place to achieve?

 

[noel_fs]

wonder how many 14700k and 14900k are gonna hit advertised 5.6hz and 6ghz now hahahahaha

 

[Broken Processor]

Seen a lot of posts on twitter/X saying they can't get chips past 5ghz with the new micro code. Will be interesting to know if these chips are already degraded or that's the norm now.

 

[ir_cow]

Here I was thinking that Microcode would patch the vcore down to like 1.4v. I've seen 1.55 before, that seems already way to high. Yet the patch limits to 1.55v.... so not a patch a all. Just Intel pretending everything is okay.

 

[zmeul]

ASUS' BIOS release for Z790 Creator is broken in some way

with Intel profile I get around 13k score in R23 (13700K) and it should've been around 29-30k

 

[Evrsr]

The problem with that is that 1.4 V is an incredibly conservative voltage and it is not enough to hit advertised clocks on at least 98% of samples, regardless of SKU type. Let's go with the 13900KS which is what I have most experience with, this is obviously a preselected, relatively high-quality i9-13900K.

The last and fastest of their kind – Intel 13th Gen Raptor Lake Binning | Part 3: i9-13900KS | igor´sLAB

Today is finally the day – loyal readers won’t be surprised by this –, for Intel’s last and fastest LGA1700 CPU we also have binning data for you once more. Thanks to MIFCOM, we were able to run…

www.igorslab.de

My personal i9-13900KS's 6 GHz V/F point is 1.463 according to my Apex Encore. and this is apparently a resoundingly average chip amongst its kind - at SP 108, 118 P and 89 E, it's smack in the middle of all bell curves from the Igor's Lab article and meets the exact median and average results of all of their samples, according to the same, for a 13th Gen i9 CPU to be considered a KS quality, its 6000 MHz point must be <= 1.49 V, so while it isn't the best, it's also far from the worst and the textbook "it is expected that it should behave exactly like this" chip.

The 13900K/KF, 14900K/KF and 14900KS processors have a different SP scale, but their V/F curve is not that far off -the i9-14900KS in particular has a 6200 MHz V/F point all the other SKUs do not - and it's common for those to be in the ~1.525 range. Now let's think outliers, such as stinker 13900K samples... yep

Sure but say 6 to 5.8GHz is a 3.4% drop, though the drop in voltage would probably be important for reliability (That article lists 1.38v for 5.8 in that particular CPU). It is probably even less at the end of the scale, when it comes to actual code that doesn't fit the caches, pushes the TLB, etc.
I imagine people would not even notice a small drop in performance and could RMA the chips if they have an especially bad CPU. I reckon 99% of the people would take that over not having their PC for a week or however long it takes to RMA the thing.

I don't have any of these but looking at benchs, even if lightly loaded, the diff from 13900K to 14900K is less than 2%. Other stuff is below 3%. I would really not risk issues or an RMA for such a small bump. If you're an overclocker, I guess these are not the chips you want.

 

[natr0n]

Saw somewhere there is billion$$$ lawsuits coming to intel.

 

[Dr. Dro]

Sure but say 6 to 5.8GHz is a 3.4% drop, though the drop in voltage would probably be important for reliability (That article lists 1.38v for 5.8 in that particular CPU). It is probably even less at the end of the scale, when it comes to actual code that doesn't fit the caches, pushes the TLB, etc.
I imagine people would not even notice a small drop in performance and could RMA the chips if they have an especially bad CPU. I reckon 99% of the people would take that over not having their PC for a week or however long it takes to RMA the thing.

I don't have any of these but looking at benchs, even if lightly loaded, the diff from 13900K to 14900K is less than 2%. Other stuff is below 3%. I would really not risk issues or an RMA for such a small bump. If you're an overclocker, I guess these are not the chips you want.

On a SP 115 chip that's the second best out of the 199 chip sample according to that data lol

 

[ShrimpBrime]

It's current limit measured in amps.

Why is everyone hung on a v-core reading.

You can pull 120 amp at 1.45v or you could pull 100 amp at 1.45v....

 

[Evrsr]

"Intel's current analysis finds there is a significant increase to the minimum operating voltage (Vmin) across multiple cores on affected processors due to elevated voltages. Elevated voltage events can accumulate over time and contribute to the increase in Vmin for the processor."

This is just confusing wording to say "high voltage causes something to happen, which, over time, makes the processor unstable at default voltage, so it would only work if we gave it higher voltage"

Yeah, this is just PR speak for degradation. Boards were requesting so much voltage this will help. Though on some chips (hopefully the good ones), the difference is minimal at best.

Don't really think this will prevent issue with the worst binned CPUs, but people are still able to set the max voltage peaks on the BIOS. That seems to be the way to go but I wouldn't want such a power hungry chip either way. Or just go with AMD or a more reasonable 13600K/14600K, if you really want a Raptor Lake die.

On a SP 115 chip that's the second best out of the 199 chip sample according to that data lol

Yeah sure, what I'm saying is that they should really set a conservative voltage and let people RMA the bad ones if they care enough.

From another Igor article, lots of 14900K samples run Fmax at 1.47v maximum, so the rest would take a small performance hit but still be way below 1.55v, which seems very excessive for long term reliability.

 

[john_]

In two weeks Intel will be OK.

YouTubers will test the patch and see minimal impact in performance, while promoting what Intel wants to be promoted. That this fix addresses the problem completely.

For the next 3-6 months any report of Intel CPU failing will be treated as a case where it was already too late to save the CPU. Intel will be accepting most RMAs and things will be looking fine.

After that period the new intel platform will be out and everyone will be looking at benchmark numbers. Even if 13th/14th series CPUs keep failing, the focus of people will be on LGA 1851 and whether 9800X3D is fast enough to help AMD to keep the performance crown in games. In apps Intel will be probably winning and that will be one more reason for people to just stop looking at 13th/14th gen fails. People reporting that theirs 13th/14th gen failed even with the fix, will start getting replies saying "Just RMA the CPU and if you don't feel fine with it, just sell it and upgrade to a different platform".

Intel is fine.

 

[Vayra86]

Highly doubt this will do anything useful. 1.55v is still a heck of a lot for this process node.

Anyone running these would be best advised to limit voltage requests in the BIOS, to at most 1.4v. Even that I wouldn't feel safe with, for CPUs running many hours per day.

1.55v is insanity on a consumer chip for the better part of the last dozen process nodes lol

Heck even 1.4v is quite the push already. Its crazy that these chips can even do this on stock. Conservative... lmao

Feels like people have been those frogs in a slowly warmed pot of water. Oblivious to the fact they're about to melt as the limits slowly got pushed up, and up.

It's current limit measured in amps.

Why is everyone hung on a v-core reading.

You can pull 120 amp at 1.45v or you could pull 100 amp at 1.45v....

Because that's also where Intel sets its limits, or used to. I remember them saying 1.5v and beyond is no go territory. And it always used to align quite well with the practical outcomes too.

In two weeks Intel will be OK.

YouTubers will test the video and see minimal impact in performance, while promoting what Intel wants to be promoted. That this fix addresses the problem completely.

For the next 3-6 months any report of Intel CPU failing will be treated as a case where it was already too late to save the CPU. Intel will be accepting most RMAs and things will be looking fine.

After that period the new intel platform will be out and everyone will be looking at benchmark numbers. Even if 13th/14th series CPUs keep failing, the focus of people will be on LGA 1851 and whether 9800X3D is fast enough to help AMD to keep the performance crown in games. In apps Intel will be probably winning and that will be one more reason for people to just stop looking at 13th/14th gen fails. People reporting that theirs 13th/14th gen failed even with the fix, will start getting replies saying "Just RMA the CPU and if you don't feel fine with it, just sell it and upgrade to a different platform".

Intel is fine.

Fine, but not great, and that's where they need to be, because their competition is.

 

[vMax65]

The problem with that is that 1.4 V is an incredibly conservative voltage and it is not enough to hit advertised clocks on at least 98% of samples, regardless of SKU type. Let's go with the 13900KS which is what I have most experience with, this is obviously a preselected, relatively high-quality i9-13900K.

The last and fastest of their kind – Intel 13th Gen Raptor Lake Binning | Part 3: i9-13900KS | igor´sLAB

Today is finally the day – loyal readers won’t be surprised by this –, for Intel’s last and fastest LGA1700 CPU we also have binning data for you once more. Thanks to MIFCOM, we were able to run…

www.igorslab.de

My personal i9-13900KS's 6 GHz V/F point is 1.463 according to my Apex Encore. and this is apparently a resoundingly average chip amongst its kind - at SP 108, 118 P and 89 E, it's smack in the middle of all bell curves from the Igor's Lab article and meets the exact median and average results of all of their samples, according to the same, for a 13th Gen i9 CPU to be considered a KS quality, its 6000 MHz point must be <= 1.49 V, so while it isn't the best, it's also far from the worst and the textbook "it is expected that it should behave exactly like this" chip.

The 13900K/KF, 14900K/KF and 14900KS processors have a different SP scale, but their V/F curve is not that far off -the i9-14900KS in particular has a 6200 MHz V/F point all the other SKUs do not - and it's common for those to be in the ~1.525 range. Now let's think outliers, such as stinker 13900K samples... yep

Not sure I understand this and maybe it is just me but I have always ignored the absurd auto voltage settings and gone for a fixed or adaptive with a 1.350v limit. My 14900K with a fixed vcore at 1.350v lets my CPU hit two caores at 6GHz with an all core 5.6GHz and totally stable. When running Cinebench R23 the voltage hits 1.360v and yet provides a 41K score and totally stable. Only when I leave vcore voltage at auto does it go into stupid territory at 1.4 to 1.5v+ during certain points but drops under load. I never lliked this and thus always kept my limit at 1.350v which actually gives me all the performance I could want...Maybe I am doing something wrong but 14900K has been runnning trouble free and that includes temps.

On the motherboard side, ASUS was the worst at pushing absurd voltages if left to it's own devices, moved to MSI Z790 Carbon Wifi which was better but still I went the fixed vcore route. Update the new BIOS yesterday and no problems whatsoever. I did contact Intel who responded fast and told me if I have any issues just RMA and within 5 to 7 days I will get a new one....With the extended warranty I will keep an eye on any issues and if they crop up, just get a new one...Intel deserve a kicking but it will come down to how the handle this for those affected and if they do the right thing and replace any 13th and 14th gen for customers then all good.

 

[Dr. Dro]

Not sure I understand this and maybe it is just me but I have always ignored the absurd auto voltage settings and gone for a fixed or adaptive with a 1.350v limit. My 14900K with a fixed vcore at 1.350v lets my CPU hit two caores at 6GHz with an all core 5.6GHz and totally stable. When running Cinebench R23 the voltage hits 1.360v and yet provides a 41K score and totally stable. Only when I leave vcore voltage at auto does it go into stupid territory at 1.4 to 1.5v+ during certain points but drops under load. I never lliked this and thus always kept my limit at 1.350v which actually gives me all the performance I could want...Maybe I am doing something wrong but 14900K has been runnning trouble free and that includes temps.

Either god sample or it's ignoring your voltage setting. More likely it's the latter.