I think you were spot on in your tuning. When I built my PC last September I set the MB to enforce the Intel limits and turn off any AI overclock. So I have not had any issues so far.

Here's the statement from Intel:

https://community.intel.com/t5/Processors/July-2024-Update-on-Instability-Reports-on-Intel-Core-13th-and/m-p/1617113

The Intel statement was dated 7/22/24. Apparently Asus issued a bios update 10 days earlier addressing a cpu voltage issue. So some motherboard makers may be ahead of the Intel on this. Suggest 13+ gen Intel cpu users check with their motherboard makers and update post-haste to the latest bios as soon as available. The 7/12/24 update from Asus is here:

https://rog.asus.com/motherboards/rog-maximus/rog-maximus-z790-hero-model/helpdesk_bios/

For what it's worth i7 are also affected. i5 also, but there don't seem to be many real world reports of issues with those. Intel says this issue doesn't affect mobile CPUs.

While the motherboard bios updates may help reduce the potential for issues by reducing the voltage the MB calls for (and default voltage and wattage levels for some were absurdly high- noticed that with my MSI board), they don't yet have the Intel microcode update for its own improper voltage levels.

There's a further problem with manufacturing defects (oxidation) affecting an unknown number of chips until some date in 2023.

Damaged chips won't be restored, so be prepared to RMA them.

Some more info putting Intel's statements into context here.

Last Christmas 2023 (*) I posted some of my experiences with my intel i9-13900K with regard of Vegas crashes due to thermal throttling. The thermal throttling itself was due to improper default CPU settings in the motherboard BIOS, driven by manufacturer competition and 'relaxed' INTEL specs and bravura.

Nowadays, I have good experience (no crashes) by setting both my power limits (TDPL1 and TDPL2) to 125W,

I lost a bit of performance but gain a lot of stability, and maybe prevent damage as the recent INTEL statements suggest updating the BIOS with new microcode (no 0x125 update yet for my mobo, I checked).

(*) original post:

https://www.vegascreativesoftware.info/us/forum/solution-to-vegas-crashes-on-modern-high-end-cpu-s-e-g-i9-13900k--144004/#ca901518

ASUS has not released a BIOS update for my motherboard yet. They are likely to start with products targeted for people trying to get the last cycle from their systems since ROG already has an update.

@bitman Controlling for TDPL1 / TDPL2 vs individual core temperatures are not mutually exclusive even if the TDPL1/TDPL2 method solves your issue on your system.

Since I built this system, I'm running much less than someone working an 8-hour day on editing.

I'm also not getting paid for doing so.

There is a lot from Gamers Nexus on the Intel issues. And a lot of the issues can not be solved with a BIOS/micro code update. Level1Techs, Wendell's take on it, especially from the server side.

Gamers Nexus is another one reporting on this, mentioning oxidation.

Apparently all the Intel 13th and 14th gen are potentially affected, 65W and up

https://www.tomshardware.com/pc-components/cpus/intel-cpu-instability-crashing-bug-includes-65w-and-higher-skus-intel-says-damage-is-irreversible-no-planned-recall

Also, based on Wendell's polling, the bios update is now considered not entirely effective by itself. Hopefully the coming microcode update to the cpu firmware will do the trick. Word from the same source is that Amd cpus, particularly the 7950X, and Intel Xeons do not exhibit the issue. Personally, I'm refraining from acquiring a 13th or 14th gen Intel cpu unless a corrective stepping is released with better thermal detection which most likely is the actual culprit.

AMD CPUs have never exhibited this issue which is a big plus!

As VEGAS Pro 21.300 and newer don't benefit as much from iGPUs, I'd have no qualms recommending or using the latest Ryzen CPUs.

AMD CPUs have never exhibited this issue which is a big plus!

@RogerS

yeah right, as if AMD does not have butter on their head, I was glad I did not have an AMD which tends to fry themselves:

https://www.forbes.com/sites/antonyleather/2023/04/30/amd-identifies-cause-of-ryzen-processors-frying-themselves/

As we've seen and read, issues do arise with just about every manufacturer. The melting plugs with Nvidia's 4090, which seems to still exist.

The difference is, AMD has fixed that issue. You can skip to the 31 minute mark. It was failure on both board partners BIOS (VSOC) and AMD in communication with them. It is no longer present, and replacement CPU's were sent out. That article is a year old. There has been no more reports. That's different with Intel and it's ongoing issue that no one knows where the cause/solution is.

Followup. Go to the 20 minute for conclusion

A BIOS update dated 07/12 appeared for my motherboard when I checked the ASUS site today. I downloaded it but I'm going to let it simmer for a while before I install it. I'll revisit my tuning strategy since it might (very likely) get reset to the factory defaults. I documented the steps with screenshots, but I'll admit that I find the whole effort tedious. I also find failure analysis tedious now that no one pays me to do it.

Ever since I built this machine, I have given some thought to rebuilding my very old media server to match my current edit workstation.

The Good

• I would have a place to send render jobs with equal performance.
• I would have a spare machine if one burns down (unless it also burns the house down).
• I would have a spare parts bin for diagnostic purposes.
• Bad publicity could depress the price of the 13900K and my overly expensive (for a media server) motherboard making it a good value for the large number of those who don't have the problem.

The Bad

• If the CPU problem is pervasive and grows over time, I could lose two machines.

The bios update will indeed reset all your settings so before doing it record any settings you want to keep and redo once updated.

An BIOS update finally appeared for my Mobo, Gigabyte Z690 AORUS MASTER, it is F29d

-Introduce the "Intel Default Settings

- Update microcode 0x125 to ensure eTVB operation normally (= the infamous intel fix 😅)

The latest on Intel here including how Intel's public statements have shifted and been quietly edited over the past year on which CPUs are affected and what customers should do:

For disclosure I am an Intel user and built a system in late 2022 with the 13600K CPU. As of yet I don't see evidence of instability with my own machine.

While I think out of memory errors in VEGAS are mostly VEGAS issues with vram, it's possible Intel CPU bugs are triggering issues here. If it affects Handbrake, Cinebench, etc. VEGAS could also be affected.

Failure rates - Puget Systems

https://www.pugetsystems.com/blog/2024/08/02/puget-systems-perspective-on-intel-cpu-instability-issues/

Apparently this can do permanent damage to your CPU, and Intel is refusing to recall the chips or replace defective ones.

The Puget one is interesting (though a fairly small sample set) as it shows that the 11th generation was by far the worst. That the 13th generation is slowly piling up failures in the field isn't great and the 14th generation has that plus more initial failures, perhaps as it's basically the 13th gen just run to its limits to perform slightly better on benchmarks and justify its existence.

Intel is replacing defective ones though has denied some RMAs and their only advice is to just try again. Extending the warranty by 2 years on their boxed products is a start.

Apparently this can do permanent damage to your CPU, and Intel is refusing to recall the chips or replace defective ones.

I believe that Intel will replace failing CPU via their RMA system. One problem is what will they replace the failing CPU with?

I have a 13900K, my brother has one and my son has one. None are showing failure yet. None have been over clocked and have had the BIOS set at Intel defaults since new.

Failure rates - Puget Systems

https://www.pugetsystems.com/blog/2024/08/02/puget-systems-perspective-on-intel-cpu-instability-issues/

Interesting read, I didn't expect to see the AMD CPU's to have twice the failure rate of Intel.