Intel processors crashing Unreal engine games (and others)

Page 83 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

DAPUNISHER

Super Moderator CPU Forum Mod and Elite Member
Super Moderator
Aug 22, 2001
30,863
28,511
146
As many of us anticipated, they are trying to bury this as fast as possible. They have not followed through on providing a testing tool -
They have not caught up with demand for tech support either - https://www.intel.com/content/www/us/en/support.html

One commenter's experience is how we got here -

If you can get through a big repack/decompressing data installation, your cpu is doing well. I had a 100% fail rate before RMA, and now a 100% success rate since RMA. When it hits all 24 cores at 100% for max power, eventually I started failing at 80% and eventually couldn't get beyond 10% without a BSOD. My CPU was so baked that at a 400a core cache current, my CPU couldn't even pull more than 160w on all core workloads.

These CPUs, while great, had more issues outside of the microcode updates. The unlimited power limits among ICC, its just crazy seeing how differently my CPU operates now vs at launch. I use to hit 100 C on a 420mm AIO, and now nothing over 75 C on the same AIO, doing the same things that use to bring it to 100 C.

Shader comp hits the CPU pretty hard too, I use to have to limit my first CPU below 160w just to finish shader comp without crashing. When they sent me a new 13900k, it's been an absolute champ.

If your CPU is rapidly degrading, you'll know it. Fastest BSOD shut downs I ever seen, sometimes the damn thing would just shut off completely.
 

DAPUNISHER

Super Moderator CPU Forum Mod and Elite Member
Super Moderator
Aug 22, 2001
30,863
28,511
146
Hmm, wonder how long that new 13900k will last?
Hopefully all of the mitigations work and there will be no need for another RMA. Even then, the CPUs being slower than the day they released is a big L in my book. Particularly seeing as at least 3 generations of Ryzen saw more significant performance uplifts than Intel from windows updates.
Regardless, it's unsurprising to see Intel moving past Raptor Lake.
Yes, the sooner it is a skeleton in the closet the happier they'll be. However, support demand says the users have not forgotten yet.

I am curious if degradation is the reason Asus nerfed this PC so radically. Steve says he bought it last summer at MicroCenter -


The bios was so old it would imply Asus knew about the problems already. The limited it to like 125W. That, or it could just be they decided to nerf it so they could sell it in a compact form factor and not have it meltdown. But typically S.I.s slap it in a cheap board with bad VRM, bad case and CPU cooling, and let it throttle. That's why I question if they knew about the problems already. The PC is stupid loud so they did not do it for acoustics.
 

Jan Olšan

Senior member
Jan 12, 2017
514
1,012
136
As many of us anticipated, they are trying to bury this as fast as possible. They have not followed through on providing a testing tool -
They have not caught up with demand for tech support either - https://www.intel.com/content/www/us/en/support.html

One commenter's experience is how we got here -
I don't remember them promising any such tool. In fact all I ever remember is reports that no, such tool is not planned. So it's silly to hold them accountable for "promise" they never made.
 

DAPUNISHER

Super Moderator CPU Forum Mod and Elite Member
Super Moderator
Aug 22, 2001
30,863
28,511
146
I don't remember them promising any such tool. In fact all I ever remember is reports that no, such tool is not planned. So it's silly to hold them accountable for "promise" they never made.
Your memory is failing you on this one. As of October Intel told Tom's they were still efforting a tool - LINK

This is the same Intel that knew about the via oxidation issues and still sent the products out. CPUs that were still on the shelves in 2024. Kept quiet and let other companies and game makers take the heat for their failures. Pointed fingers at their board partners. Early on denied RMAs in masse to S.I.s. and many end users. Released information through community posts; laughable. What's silly to me, is defending Intel over anything to do with this historically FUBAR epoch.
 

DAPUNISHER

Super Moderator CPU Forum Mod and Elite Member
Super Moderator
Aug 22, 2001
30,863
28,511
146
GN revisits the epic fails of last year, and this debacle was predictably and humorously included. The whole vid is high entertainment but vid is time stamped for pertinent section -

 

Kocicak

Golden Member
Jan 17, 2019
1,177
1,232
136
They have not followed through on providing a testing tool.
Testing tool? What should it test? How many combinations of motherboards, bioses and settings are possible? I cannot imagine any voltage testing would bring reliable results in such situation.

For testing stability of the CPUs users already have the applications they need to run and other stability testing tools available. There is an official Intel CPU stability testing tool, which runs a few short loads - how could that be better than what people already have?

There is a problem here: extensive stability testing itself can lead to CPU degradation. Somebody can have a CPU which could last him a few years with his use case with no problem, but he could substantially shorten its life if he tested it for stability too much.
 
Last edited:

DAPUNISHER

Super Moderator CPU Forum Mod and Elite Member
Super Moderator
Aug 22, 2001
30,863
28,511
146
Testing tool? What should it test? How many combinations of motherboards, bioses and settings are possible? I cannot imagine any voltage testing would bring reliable results in such situation.
Don't ask me these questions, ask Intel. It's their BBQ that's burnt, not mine.
For testing stability of the CPUs the users already have the applications they need to run and other stability testing tools availale. There is an official Intel CPU stability testing tool, which runs a few short loads - how could that be better than what people already have?
See above.
There is a problem here: extensive stability testing itself can lead to CPU degradation. Somebody can have a CPU which could last him a few years with his use case with no problem, but he could substantially shorten its life if he tested it for stability too much.
That certainly seems to be the case with rapture lake lol. I have serviced 15yr old daily drivers where components failed, but the CPU was chugging right along.
 

IEC

Elite Member
Super Moderator
Jun 10, 2004
14,565
5,961
136
My 14700K has been flogging the 1T/2T boost running F@H since early December. I've also been doing some light UE4-based gaming on it which is another not-so-great for degradation scenario.

Has been run on Microcode 0x12B since I got it (was flashed on a 12600K prior to installing).

So far, it's okay. No BSOD or reboots, no loss of boost clocks. Will keep reporting back since obviously degradation would take time and this should represent a near worst-case scenario for Raptor Lake CPUs (lightly-threaded, high boost).
 

Kocicak

Golden Member
Jan 17, 2019
1,177
1,232
136
That certainly seems to be the case with rapture lake lol. I have serviced 15yr old daily drivers where components failed, but the CPU was chugging right along.
Are you sure that even the modern TSMC processes are so robust and long lasting as the 15 year old tech used to be? It would be nice if somebody could find a compaison of size of the transistors in CPUs of today and past.
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
26,820
15,821
136
Testing tool? What should it test? How many combinations of motherboards, bioses and settings are possible? I cannot imagine any voltage testing would bring reliable results in such situation.

For testing stability of the CPUs users already have the applications they need to run and other stability testing tools available. There is an official Intel CPU stability testing tool, which runs a few short loads - how could that be better than what people already have?

There is a problem here: extensive stability testing itself can lead to CPU degradation. Somebody can have a CPU which could last him a few years with his use case with no problem, but he could substantially shorten its life if he tested it for stability too much.
There is a major problem with a lot of this logic. Read below.

First, the vast majority of CPUs is not to DIY people, it is OEM, this is a FACT.
Second, lets take a few cases : OEM from Best buy, or some other retail store: A customer has a problem with one of the affected models. They have no idea whats wrong or how to spell Intel. They take it back to the store. How knowledgeable is the technician there ? How long is the customer inconvenienced ? Based on a lot of factors, how much does it cost the customer in lost time , and based on other factors like is the CPU RMA granted, is it cross shipped ? Does the tech even know about this specific problem ? end result: NOt good for the customer and is NOT ACCEPTABLE for a product to do this.

Now lets take a corportorate case, and numerically most likely the highest number of cases. The PC has a problem. The local IT staff sees the problem. How do they address it ? Do they have spares ? Does Intel grant the RMA ? What function was the computer performing and how much loss is there in many ways to the company ?

In every case there is a loss to the user, some times more than others, and in different ways. Is it acceptable to sell a product like this ? Absolutely NOT ! Does is cost the end user money in one way or another ? YES. How is the situation handled by the owner ?

This is totally unacceptable for a company to produce a product like this, let alone how they have handled it. There is no excuse from any angle as to how this happened or how it was handled by Intel.

And lastly, how much damage has the high electricity usage of this line of processors cost the environment ?

Way too many questions, but the bottom line is Intel made a bad product and handled the case badly.
 

Kocicak

Golden Member
Jan 17, 2019
1,177
1,232
136
.....

Way too many questions, but the bottom line is Intel made a bad product and handled the case badly.
That is true.

The other point of view is that now you can get 20 or 24 core CPUs for very good price, which will serve you well, once you understand their limitation and set the power draw and frequency limits so that the CPUs cannot damage themselves.

So for knowledgeable people these CPUs are a great deal.
 
Reactions: igor_kavinski

APU_Fusion

Golden Member
Dec 16, 2013
1,589
2,367
136
That is true.

The other point of view is that now you can get 20 or 24 core CPUs for very good price, which will serve you well, once you understand their limitation and set the power draw and frequency limits so that the CPUs cannot damage themselves.

So for knowledgeable people these CPUs are a great deal.
Why should it be the responsibility of the end-user of their product to do that? My 4790k is happily running internet and office apps.
 

Kocicak

Golden Member
Jan 17, 2019
1,177
1,232
136
Why should it be the responsibility of the end-user of their product to do that?
It should not be. But in spite of what should or should not be, the situation now is that the CPUs are still sold with those mad settings. If you buy a new 13900KS today, it will still default to 320W power limit. That is insane.

If you know what is happening and you can make a few basic settings in BIOS, you can get a great value product. 14700K is a 20 core CPU and the closest AMD price competitor from AMD has 8 cores.
 
Last edited:

APU_Fusion

Golden Member
Dec 16, 2013
1,589
2,367
136
It should not be. But in spite of what should or shouild not be, the situation now is that the CPUs are still sold with those mad settings. If you buy a new 13900KS today, it will still default to 320W power limit. That is insane.

If you know what is happening and you can make a few basic settings in BIOS, you can get a great value product.
Again, why should this be the responsibility of the end-user? You seem to be arguing this isn’t a major issue of defective cpus being knowingly sold because a tiny fraction of people know enough about computers to “maybe” mitigate the issued. The 13900k and 14900k should never been released it seems. I find it odd to state these are good values over AMD even if you know how to mitigate the issues wither failure rate.
 

In2Photos

Platinum Member
Mar 21, 2007
2,402
2,624
136
That is true.

The other point of view is that now you can get 20 or 24 core CPUs for very good price, which will serve you well, once you understand their limitation and set the power draw and frequency limits so that the CPUs cannot damage themselves.

So for knowledgeable people these CPUs are a great deal.
A great deal everyone! I think not. Only way I would EVER consider using one of these chips is if you sent me the rest of the parts for the build and I used it for something I don't need to rely on.
It should not be. But in spite of what should or should not be, the situation now is that the CPUs are still sold with those mad settings. If you buy a new 13900KS today, it will still default to 320W power limit. That is insane.

If you know what is happening and you can make a few basic settings in BIOS, you can get a great value product. 14700K is a 20 core CPU and the closest AMD price competitor from AMD has 8 cores.
Running a 14700K at reduced settings is probably still slower than the AMD 8 core version. And it's end of life. Still don't want it. Quit trying to justify these chips. They will likely go down as one of the worst CPUs of all time.
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
26,820
15,821
136
That is true.

The other point of view is that now you can get 20 or 24 core CPUs for very good price, which will serve you well, once you understand their limitation and set the power draw and frequency limits so that the CPUs cannot damage themselves.

So for knowledgeable people these CPUs are a great deal.
Buying a CPU that could STILL degrade and reboot is not a great deal. Not to mention, you have to turn down the wattage a lot, and then the performance BITES.

You should stop defending a dead horse.
 

Thibsie

Golden Member
Apr 25, 2017
1,010
1,186
136
It should not be. But in spite of what should or should not be, the situation now is that the CPUs are still sold with those mad settings. If you buy a new 13900KS today, it will still default to 320W power limit. That is insane.

If you know what is happening and you can make a few basic settings in BIOS, you can get a great value product. 14700K is a 20 core CPU and the closest AMD price competitor from AMD has 8 cores.

You meant a very BAD deal, didn't you ?
Even for free, I wouldn't touch it.
 

DAPUNISHER

Super Moderator CPU Forum Mod and Elite Member
Super Moderator
Aug 22, 2001
30,863
28,511
146
If you ever doubted it's Stockholm Syndrome -


The 2nd one died despite an up to date bios. This is the first report I've read of that happening. EDIT: I need to RIF =Reading is fundamental It hasn't been long since the mitigations were finalized. Going to be interesting to see what reports are like by this time next year. If users bother to provide feedback the way Hulk did.
Even for free, I wouldn't touch it.
If they are going to keep dying despite the mitigations, that's the only way I'd take one. It'd be fun to use it on games with long shader compiles or lots of asset streaming like the Spiderman games and see if it 💩 the bed. And if so, how long it takes. Because I can go 6-8hrs straight on a game; sometime longer. Last of Us 4K max settings had my PC throwing out the warmest air I've ever felt from it after a session that long. 7800X3D+7900XTX runs perfectly of course.
 
Last edited:

DAPUNISHER

Super Moderator CPU Forum Mod and Elite Member
Super Moderator
Aug 22, 2001
30,863
28,511
146
Read that again. The replacement for the first CPU came BEFORE the BIOS update and has been used heavily before the BIOS update.


Speaking of Stockholm Syndrome...

My comment was about the post mitigations failure. Had you read my post correctly.

"Unfortunately, I started experiencing stability issues a bit recently despite update-to-date bios which prompted a second ticket. Intel both times has responded well"

has been used heavily

This has been a drum you keep beating. If I can't use my CPU heavily for fear of failure, it's broken.
 

DrMrLordX

Lifer
Apr 27, 2000
22,491
12,365
136

Speaking of Stockholm Syndrome...

My comment was about the post mitigations failure. Had you read my post correctly.

"Unfortunately, I started experiencing stability issues a bit recently despite update-to-date bios which prompted a second ticket. Intel both times has responded well"



This has been a drum you keep beating. If I can't use my CPU heavily for fear of failure, it's broken.
Technically I think he's correct. The reddit user had a bum 13700k, got a replacement, ran it before updating his UEFI, scorched it too, and got a replacement 14700k because Intel has probably run out of 13-series chips and has no choice in the matter.

At least this confirms that updating the UEFI while using a Raptor Lake that has operated "heavily" under an older microcode version is not a viable fix.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |