Intel processors crashing Unreal engine games (and others)

[SamMaster]

That is good to read.

Yes, thankfully we have an authorized service provider that makes things easy for us (except for Dell, they make things difficult across the line).

 

[DAPUNISHER]

Oh boy! Mike Goldberg - HERE....WE....GO!

https://www.reddit.com/r/pcmasterrace/comments/1ikjnpi/intel_13th14th_gen_update_0x12b_doesnt_seem_to/

 

[igor_kavinski]

My favorite:

 

[Markfw]

My favorite:

View attachment 116600

I agree with MarcCDB, there is no reason for anybody to buy any of the 13th or 14th gen Intel CPU's now that we have seen many reports of them dying at the mid and lower range levels as well as 1x900k.

I personally don't trust the 12xxxk series as well, I would never buy any of them. And anything old than that do not perform anywhere close to their AMD counterpart. The newest gen 2x5k's not only don't have avx-512 support, but have odd performance problems, so those are out as well. And its not just performance, AMD's efficiency is better as well. Many places a LOT more efficient.

To sum it up, would not buy any CPUs older than 2017 for performance reasons, and no Intel for 2017 or newer for the reasons I stated above.

 

[igor_kavinski]

I personally don't trust the 12xxxk series as well

I have tortured my 12700K a lot using Intel XTU. Still alive. Though I could've sworn that I could overclock it higher before. Maybe the socket got warped. Now that I have a 245KF, I wish I could use the 12700K as my work PC. Does feel a lot faster than any of the PCs that I've used in my office. I would've gone with 9950X3D if AMD had done it like before when they launched the 7950X3D first. Unfortunately, I couldn't wait anymore (winter is the best time for overclocking experiments for me. It's quiet without the sound of airconditioner running).

 

[Markfw]

I have tortured my 12700K a lot using Intel XTU. Still alive. Though I could've sworn that I could overclock it higher before. Maybe the socket got warped. Now that I have a 245KF, I wish I could use the 12700K as my work PC. Does feel a lot faster than any of the PCs that I've used in my office. I would've gone with 9950X3D if AMD had done it like before when they launched the 7950X3D first. Unfortunately, I couldn't wait anymore (winter is the best time for overclocking experiments for me. It's quiet without the sound of airconditioner running).

I did say "personally" as I don't have any real evidence against this CPU or below in the Alder lake series.

 

[Hans Gruber]

PUBG is completely unplayable on my 7600K. It used to work and have weird artificing. In the past Intel CPU's ran every game without issue. Intel must have done something to their microcode that has screwed up all games based on the unreal engine. PUBG is not a new game and used to run like butter on Intel systems.

 

[reb0rn]

I will be honest, I have one 13700k works OK mean nothing, had 20+ AMD 5x and 7x and now 9x, they had a lot issue, maybe mu luck but 2-3 are semi defected... and yeah its quite hard to prove CPU issue when is just unstable

This topic sadly do not tell me anything, just bashing what who likes more.......

Defo intel gone over the top with 13 and 14gen power use but AMD did same but AMD is now geek temple of love and no one can say anything bad for someone to believe it, I see quite a few post all both side for new CPU and issue, I guess intel could be worst by pure speculation and that is

Rest is FUD

 

[DZero]

My favorite:

View attachment 116600

Heck, even Intel 12th gen is far better now! Even the Xeons!

 

[Thunder 57]

I will be honest, I have one 13700k works OK mean nothing, had 20+ AMD 5x and 7x and now 9x, they had a lot issue, maybe mu luck but 2-3 are semi defected... and yeah its quite hard to prove CPU issue when is just unstable

This topic sadly do not tell me anything, just bashing what who likes more.......

Defo intel gone over the top with 13 and 14gen power use but AMD did same but AMD is now geek temple of love and no one can say anything bad for someone to believe it, I see quite a few post all both side for new CPU and issue, I guess intel could be worst by pure speculation and that is

Rest is FUD

Lol. Anecdotes are not facts. It is a fact that 13 and 14th gen degrade. Maybe it's just user error? Or maybe you are spreading FUD.

 

[DAPUNISHER]

I will be honest, I have one 13700k works OK mean nothing, had 20+ AMD 5x and 7x and now 9x, they had a lot issue, maybe mu luck but 2-3 are semi defected... and yeah its quite hard to prove CPU issue when is just unstable

This topic sadly do not tell me anything, just bashing what who likes more.......

This thread is pages and pages of hard evidence. Intel has acknowledged everything. It has become an expensive problem and the costs are ongoing.

Rest is FUD

The only FUD is coming from you. Any further nonsense will be deleted. Gaslighting and astroturfing do not serve the public interest/help affected users understand the causes of instability and seek remediation, compensation, etc.

 

[DrMrLordX]

Oh boy! Mike Goldberg - HERE....WE....GO!

https://www.reddit.com/r/pcmasterrace/comments/1ikjnpi/intel_13th14th_gen_update_0x12b_doesnt_seem_to/

Read this comment from the reddit post:

Even with power limits or offset undervolting, the VID request could still be above 1.5V at times, even if it was rare. There was a bug with the algorithm, single core boost requesting high amounts of voltage to hit the advertised clockspeed, vendors motherboards doing their own spin on things (usually overvolting to get more performance). Absolute disaster from all fronts and the hardware scene needs to take a step back to prevent anything like that happening ever again (but it won't).
Personally I always tinker with hardware. I go the more efficient approach. I messed with LLC and manual voltage. Ended up going with flatline 1.27V for 5.6Ghz (not the greatest bin but I don't care too much about +100-200Mhz)

The chip in question was run with old microcode for quite some time. We'd need to see a relatively-new 14900k degrading under the new microcode for it to be a major scandal. That being said, anyone who thought that they were safe from using moderate power limits under the old microcode probably wasn't. It seems like anything higher than 180W PL2 and/or a significant undervolt was unsafe.

 

[DAPUNISHER]

The reason I wrote it that way, is the trolls have been pushing back hard lately. Not because I thought it is a new CPU post mitigations that failed. It makes no sense, but they have been downvoting and trolling anyone reporting a broken raptor lake CPU lately. With Intel issuing the mea culpa of sorts and releasing a handful of fixes, it is now simply a matter of record that all 65W and above 13 and 14 gen in use prior to the final fixes, are susceptible to accelerated degradation damage. What the trolls hope to accomplish by denying reality is beyond me. It is a really bad look.

 

[DAPUNISHER]

Here's a prime example from that thread -

HURR DURR! "How "degrade" only affect UE5 engine and not other apps or BSODs? Maybe problem is UE engine themself?" HERP DERP!

 

[DrMrLordX]

The reason I wrote it that way, is the trolls have been pushing back hard lately. Not because I thought it is a new CPU post mitigations that failed. It makes no sense, but they have been downvoting and trolling anyone reporting a broken raptor lake CPU lately. With Intel issuing the mea culpa of sorts and releasing a handful of fixes, it is now simply a matter of record that all 65W and above 13 and 14 gen in use prior to the final fixes, are susceptible to accelerated degradation damage. What the trolls hope to accomplish by denying reality is beyond me. It is a really bad look.

Okay, that's fair.

 

[DAPUNISHER]

My favorite:

View attachment 116600

How about this

 

[511]

How about this

Microcode would not fix broken CPU someone needs to buy a new one for science and torture it for Science and let us know the results one year later for science.
12th gen they is fine and there is no issues buying rebranded ADL CPUs as well it is the 8+16 Die causing issues thanks to the guy who forgot to work correctly on clock tree.

 

[DZero]

Microcode would not fix broken CPU someone needs to buy a new one for science and torture it for Science and let us know the results one year later for science.
12th gen they is fine and there is no issues buying rebranded ADL CPUs as well it is the 8+16 Die causing issues thanks to the guy who forgot to work correctly on clock tree.

Indeed, but even if at the end the microcode works, the image of the 13th and 14th gen are beyond repair.
They ruined a decent uArch which is Alder lake. I can't believe how that was forgotten
What are the quality assurance with that?

 

[igor_kavinski]

Microcode would not fix broken CPU someone needs to buy a new one for science and torture it for Science and let us know the results one year later for science.

I don't think that a sample size of one is going to matter. The comments have some people running their 13th gen CPUs without the microcode update and even factory BIOS and no degradation for them. It seems to be a quality control issue where only a fraction of the CPUs are susceptible to degradation but that fraction could be anywhere from 5 to 50% in a single batch. This is what Intel isn't telling anyone and no one seems to be leaking it from the front lines either.

 

[DAPUNISHER]

I don't think that a sample size of one is going to matter. The comments have some people running their 13th gen CPUs without the microcode update and even factory BIOS and no degradation for them

I trust them far less than users having issues. If they do have degradation, they are not going to tell you about it. Amygdala is too big, resulting in feels being too strong.

Disclaimer: The following is all from memory so if I get minor details wrong, that's on me.

Steve of Random Gaming in HD, I am confident has a broken 14900K. He used it for content after Intel gave it to him through a UK retailer. Then when it reappeared later, for some undeclared reason, it was no longer paired with a 380 or 420mm AIO, but rather on a cheap air cooler running at a fixed 125W. He removed CPU info from afterburner when using it. Now he uses a 12700kf as his regular test rig, and the AB info is back. 5 will get you 10 his 14900K degraded.

I have also seen 2 of the owners of Star Forge PC have their PCs fail during streams. What do you want to wager they were both raptor. Asmongold kept recompiling the shaders every time it crashed until it made it in game. It took like 50 minutes. They were guessing how far it would make it each time before crashing. I think I posted that some pages back? Not the best advertisements for their PC company to have their expensive custom PCs fail trying to play Space Marine 2 and Marvel Rivals IIRC.

. It seems to be a quality control issue where only a fraction of the CPUs are susceptible to degradation but that fraction could be anywhere from 5 to 50% in a single batch. This is what Intel isn't telling anyone and no one seems to be leaking it from the front lines either.

What workloads and amount of uptime are involved is what seems to dictate failure. As this latest user noted, all of the manual settings being hailed as how to avoid the issue early on, can and do fail. Minecraft and Path of Titans servers will kill every single raptor from what I've seen. No silicon lottery or conservative board will save them.

 

[511]

PugetBench applied custom settings and they somehow were not affected much it's ridiculous a third party know how to set correct processors settings than major Intel OEMS

 

[Ranulf]

Here's a prime example from that thread -

HURR DURR! "How "degrade" only affect UE5 engine and not other apps or BSODs? Maybe problem is UE engine themself?" HERP DERP!

It is funny and frustrating to watch the PR firms work their hilariously bad magic at times. To be fair to the reputation rejuvenator monkies I'm not a big fan of UE5 games these days but I also don't see shader compiler crashes on a 10 year old 4th gen i7. Other game crashes yes (hello Marvel Rivals) and those are affecting many across platforms. I've also been watching a streamer lately that for whatever reasons refuses to get his 13900k replaced and will have random crashes in various games. I think he finally listened to his mods and chat to downclock the thing at least.

 

[Thibsie]

PugetBench applied custom settings and they somehow were not affected much it's ridiculous a third party know how to set correct processors settings than major Intel OEMS

PugetBench... Smells like DellBench to me.
Not even a slight bias, just a huge one. IMO.

 

[DAPUNISHER]

PugetBench... Smells like DellBench to me.
Not even a slight bias, just a huge one. IMO.

Agreed. He is shilling and lying in his own interests.

He represents his company on Intel's board of advisors. He is offering among the shortest extended warranty periods of 3 years for affected systems. If his failure rates are so low why not 5 years like most other S.I.s? Because he is lying.

Suddenly AMD is the most reliable and Intel 13 and 14 have stability problems? You don't say? - https://www.pugetsystems.com/labs/a...most-reliable-hardware-of-2024/#CPU_Processor

Now AMD is the top selling lineup for him. Is that because you recommend them, or customers demand them? I surmise both. He ran cover to protect his earnings, but you can bet they were steering customers away from raptor lake at the same time he was making those BS claims.

 

[IEC]

Microcode would not fix broken CPU someone needs to buy a new one for science and torture it for Science and let us know the results one year later for science.
12th gen they is fine and there is no issues buying rebranded ADL CPUs as well it is the 8+16 Die causing issues thanks to the guy who forgot to work correctly on clock tree.

See my post earlier in the thread

I updated to 0x12B on a 12600K and swapped in a retail 14700K. It's running F@H on the GPU (1t max boost load most of the time) and otherwise being used for lightly-threaded gaming with things that flog 1-2 cores. I purchased the 14700K in early December so it's been about 2 months without issue so far. Only 10 more months to go...

BIOS7D27v1J (0x12B) on MSI Z690 Ace