Intel has a Pretty Big Problem..

[Rroff]

* 3% - pretty bad but not end of world)

Just for context I used to be involved with running a GSP for several years with lots of dedicated servers - having an Intel CPU failure at all was almost unheard of, and usually if there was a failure it was DOA if it did happen - it was usually something else which failed (if anything failed at all), so even 3% would be quite an abnormal figure.

 

[Tetras]

Any replacements not being an actual random sample is certainly one way that figures like 3%*, 20%**, and 50%*** could be reconciled.

Problem I see with that theory is, how many non-random samples from each tray batch would Intel support hold?

Only really makes sense if 1st CPU goes, 2nd (from a different batch) also goes bad, 3rd (this time from the same batch as the 2nd) goes bad. That way Intel support only have to send without testing and/or have no clue about how batches tie in with the problem.

* 3% - pretty bad but not end of world)
** 20% - very very bad
*** 50% - wind up the company bad

I think it would work as a theory only if Intel are reserving very large numbers of batches of tray CPUs for specific OEMs.

They then send those replacement CPUs from the reserved set (to account for failures) and from that set, a large number of those batches which are produced between maybe 2-3 months are contaminated at least to a partial degree - since if it applied to just a few batches and just from one particular location, it couldn't possibly account for the number of faulty CPUs reported, or Intel's apparent inability to pin it down to and pull out of circulation specific CPUs.

 

[TrixP10]

Try not to worry. Enjoy the laptop and cross fingers for now. It may be worry over nothing.

I wouldn't risk it. would you?

 

[thebennyboy]

I do wonder whether this could potentially affect lower-end SKUs eventually. Could power and heat just be expediting the failures?

 

[Journey]

I'm eyeing my new laptop and worrying. It's not like they can replace the chip, like a desktop.

Wait, or get something with a Ryzen in it...

 

[eeii]

Just for context I used to be involved with running a GSP for several years with lots of dedicated servers - having an Intel CPU failure at all was almost unheard of, and usually if there was a failure it was DOA if it did happen - it was usually something else which failed (if anything failed at all), so even 3% would be quite an abnormal figure.

Have to agree with this. I did server support for over 20 years. In that time I don’t think we had any cpu failures. Thermal paste needing replacing was about it.

 

[Tetras]

I do wonder whether this could potentially affect lower-end SKUs eventually. Could power and heat just be expediting the failures?

Most of the lower SKU CPUs use the 12th gen die (C0), so apart from the 14600 and the 13400/14400 (they're C0 or B0) I don't think it'll apply to them. The only posts I've seen with lower SKUs reporting problems they had a CPU with a B0 die.

 

[q974739]

Wait, or get something with a Ryzen in it...

It's too late. Bought in January. It's the ticking risk though?

 

[Gorbash12346]

Potentially joining the boat of 13900K starting to fail. On Intel's performance setting in my BIOS, rolled back to 4800 on my ddr5 and now running PCIE 3.0 just to stop it from crashing now. Going to start the ball rolling on trying to get a warranty claim on Monday

 

[ED209]

Potentially joining the boat of 13900K starting to fail. On Intel's performance setting in my BIOS, rolled back to 4800 on my ddr5 and now running PCIE 3.0 just to stop it from crashing now. Going to start the ball rolling on trying to get a warranty claim on Monday

not good at all, how long have you had it for?

 

[Gorbash12346]

not good at all, how long have you had it for?

Built December 2022.

 

[KompuKare]

Just for context I used to be involved with running a GSP for several years with lots of dedicated servers - having an Intel CPU failure at all was almost unheard of, and usually if there was a failure it was DOA if it did happen - it was usually something else which failed (if anything failed at all), so even 3% would be quite an abnormal figure.

Yes, I think Intel's quoted figures (in the GN video, "... one source told us that Intel is telling OEMs that it has observed a 0.035% worldwide failure rate ...") is about what I would expect from a tier 1 silicon company - anything higher would imply poor QA and binning. And I would also expect 80-90% of 0.035% to be DOA rather than failing later.

3% would not kill Intel but is around x100 higher than expected. The other figures are almost unheard of (Nvidia's bumpgate parts had near 100% in that they should all eventually fail but some less thermally stressed parts might still be alive; but then that was packaging, poor choice of underfill rather than the leadfree solder as such).

 

[KompuKare]

It's too late. Bought in January. It's the ticking risk though?

You might be lucky. While exact cause is unknown and CPUs on the server/workstation W680 boards also failing is worrisome, thermal stress or excess voltages might still play a role. What laptop is it? I would suspect a lower wattage part might be safe enough - at least for a time - but desktop-replacement parts? Who knows

Mind you, with RPL's power usage, did Intel actually make sense in any recent laptop?

Potentially joining the boat of 13900K starting to fail. On Intel's performance setting in my BIOS, rolled back to 4800 on my ddr5 and now running PCIE 3.0 just to stop it from crashing now. Going to start the ball rolling on trying to get a warranty claim on Monday

Do tell is your RMA experience. All smoothly is what we would expect of Intel CS. At least Intel CS of yore, anything less we would certainly like to know about.




About the AMD überfans quip -while I am very glad to not have to give Intel my money (at least until AMD catch up on the nasty shenanigans Intel have done throughout the decades), my main worry with Intel was the power usage.

I already though the whole pre-overclocked out the factory for CPUs and GPUs the last decade or so was very bad. But to then see Intel pushing CPUs which can consume over 300W - and that on the desktop mainstream platform. Well that is something I did not like one bit. Especially since those i9's sold quite well as far too many bought them as they were 1-2% quicker than a Ryzen 3D at sometimes a quarter of the power (although for gaming those i9's were often nowhere near 300W their FPS/W was still awful).

Same with 4090 and the rumoured +50W TDP for the 5090.

Yes, I know nodes are getting more expensive so the temptation is to at the very edge stability (or possible beyond...), but these kind of things often gain no more than a few %. Save 30% of power, lose 1-2% is quite common.

What I'm trying to say is that if crazy power and clocks ends up being a factor in this and it costs Intel a lot of money and goodwill: well I will be glad and that's not even Schadenfreude just hopefully something encourages all the PC companies to return to some sanity. Leave a few % on the table and let extreme overclockers play with them, not everyone.

 

[sx_turbo]

Remember when the amd 7k chips fried themselves and was only a small handful of people it happened to and the world and media outlets lost their minds, AMD had a fix pretty swiftly within a few weeks I seem to remember.

As has been said in the first post, seems to be much quieter from the news outlets now it's intel. Where's outraged Steve and jayzee when I need them...

 

[ED209]

Built December 2022.

Only good thing is its under warranty and hopefully your replacement will be fine.

Good luck with the RMA Monday and as @KompuKare said please let us know how the rma goes and if you have any issues.

Are you doing the RMA directly with Intel or the reseller?

 

[Gerard]

Remember when the amd 7k chips fried themselves and was only a small handful of people it happened to and the world and media outlets lost their minds, AMD had a fix pretty swiftly within a few weeks I seem to remember.

As has been said in the first post, seems to be much quieter from the news outlets now it's intel. Where's outraged Steve and jayzee when I need them...

Do you recall those 6000 gpus that fried themselves last year or thereabouts? The media were losing their **** about it, blaming drivers etc. Yet for some reason they didn't seem to clue into the fact it was like 60 gpus in a single country, and it later turned out they were used for mining, likely soaked to clean them then stored in a damp environment.

Funny how that grabbed so many headlines jumping on the "drivers to blame" bandwagon, musta been the most selective driver bug in history if it had turned out to be driver related, only affected a specific amount of gpus in Germany, yet that seemed to be lost on all the media outlets who were trampling over each other to be the first to put up a YouTube video with a really suspect headline about it.

 

[Vexr]

Frame Chasers put out a intresting video yesterday which I can't link here as it has swearing in it though Jufus says that the single core/two preferred cores boosting is giving the CPU too much voltage is the issue thats causing the CPU's to degrade and locking the cores to stop this boosting should stop it from happening.

The video also goes into detail of an degraded 13900KS he had recieved from a discord member that crashes after loading bios optimised defaults and put under a little load/pretty much idle in Windows. To make the CPU stable one has to lock the cores and up the voltage to make the CPU stable to account for the degradation that has already happened though some people in the comments said this needs more extensive testing.

This could explain why even with power limits in place this would still happen though its still not clear if this is an issue from the factory or a motherboard issue or a combination off both and/or other issues.

I've dropped this comment on a video of his, on the basis I'm someone who has watched his tripe for a while now:

You're the ultimate insane Intel fanboi. I recall when you were saying something along the lines of 'if you're a serious performance CPU buyer, then you'll be performing hardware surgery delidding your mega-expensive Intel processor with a razor blade, invalidating any warranty, and tweaking this and that in the BIOS for your 360mm AIO to just about be able to cope.

Also recall you ranting about the 'AMDrop' even on an X3D that you could only replicate spinning around in some donkey-ass game like ARMA 3 or some BS

What ratio of your strategy is just gaslighting vs you being genuinely insane, I don't give a damn, but suspect it's the latter + money that has made you this way. Apparently people pay you for 'consultations' (bet that's a laugh for the customer) / throw money at you because they're almost equally mad or desperate Intel fanbois

Also "HUB is the devil for not putting an extreme RAM OC in their Intel reviews" or some BS

You're insane.

 

[MrMarvelous]

Could we keep a track of how many people here claim on warranty and how many are successful? Unfortunate for owners but I think the data would be interesting.

 

[jonneymendoza]

Intel are done

 

[Gerard]

Intel are done

AMD were also done at one point, intel are nowhere near that, and with competition being good for us hopefully they come out with some good products in the future.