Intel 13th and 14th Gen confirmed defective

[SpyderTracks]

Uh oh, reports coming in about 13th and 14th gen mobile cpus starting to get odd crashes

Intel says 13th and 14th Gen mobile CPUs are crashing, but not due to the same bug as desktop chips — chipmaker blames common software and hardware issues

Not the same instability that's affecting desktop Raptor Lake and Raptor Lake Refresh processors.

www.tomshardware.com

Intel laptop CPUs are crashing, too, as the company remains silent

Shortly after blasting Intel for the high failure rate of the company's recent CPUs, Alderon Games claims laptops running on Raptor Lake chips are no exception. The...

www.techspot.com

Intel are saying it’s not related to the desktop cpu issues, but the community feel differently

Dev reports Intel's laptop CPUs are also suffering from crashing issues — several laptops have suffered similar failures in testing

New evidence busts any theories that Raptor Lake instability is a desktop-specific issue.

www.tomshardware.com

 

[Scott]

I was wondering about the laptops. If it's gone that far then I don't know how Intel are going to come through this. Every single sector will be affected.

 

[SpyderTracks]

I was wondering about the laptops. If it's gone that far then I don't know how Intel are going to come through this. Every single sector will be affected.

yeah, 100%. If this gets confirmed to be degradation on the mobile cpus I expect regulators to step in, Intel just seem adamant to gaslight at every stage. I’m betting they already know the scale of this, just aren’t admitting it.

 

[Scott]

I would be absolutely flabbergasted if they don't already know. They will have already done the tests that GN is out-sourcing so I would imagine they already have the root cause at hand and know the spread of the issue already. I'm not surprised there hasn't been a disclosure of this but my guess is the lawyers and PR teams are getting overtime.

 

[SpyderTracks]

Skip to 44:48

this is MooresLawIsDead, this guy is HUGELY connected in the CPU industry and often leaks things well in advance of other sources.

The word Recall is starting to be floated for Raptor Lake

 

[SpyderTracks]

Hmmmm, Intel have an official response finally!

TLDR: They're saying it's too high voltages and are rolling out a microcode update to set proper limits

Intel Publishes Official Response To 14th Gen & 13th Gen CPU Instability Issues: Cites Elevated Operating Voltages As Root Cause, New Microcode Patch In August

Intel has an official statement regarding 13th and 14th Gen CPU instability issues, citing elevating operating voltages as the root cause.

wccftech.com

Expected release is Mid-August

I'm calling it right now, this is only part of the problem from what we've been told by several experts who've done extensive analysis on this.

A lot of server side installations in datacenters are experiencing the same issues, and those motherboards DO NOT ALLOW over voltage. This would also suggest it's only affecting K series CPU's which can be unlocked to draw extra voltages, but it's been confirmed to be affecting non-k cpu's also.

1/. Unlimited voltages allowed IMPLICITLY confirmed as "In Spec" by Intel, they still haven't stated otherwise, although are now rolling out this patch to limit

2/. Oxidisation on the VIAs, due to incorrect application of anti-oxidant at the factory. This results in higher resistance which means even if you are passing through lower voltages, the resistance is going to result in increased thermals in those sections

3/. It's been confirmed by consumer and professional sources that a lot of CPU's out of the box on boards with full power limits in place are still experiencing instability.

This is by no means fixed, and quite frankly it's insulting from Intel just further ignoring the communities feedback.

GN in the video above have employed a Failure Analysis lab to basically tear down several affected CPU's to determine at a microscopic level what can be seen from oxidisation through to silicon deterioration, and which specific areas may be affected.

This is not done in any way, Intel should have made that clear in this statement! They've stated this is a root cause fix which means they're closing the book, totally unacceptable.

Ok, so there is a note about the Oxidisation issue, I totally missed it as it’s at the bottom of the article on Toms Hardware. I’m unsure why this wasn’t in the other article, I would have thought intel would have given one press announcement to all partners

Intel finally announces a solution for CPU crashing and instability problems — claims elevated voltages are the root cause; patch coming by mid-August [Updated]

Patch is on the way.

www.tomshardware.com

”

Intel statement on via oxidation​

Short answer: We can confirm there was a via Oxidation manufacturing issue (addressed back in 2023) but it is not related to the instability issue.
”

This still strikes me as odd, if they know when the manufacturing process was improved, then they know specifically which batches are affected.

Normally that would trigger a recall of the affected batches. If we’re only starting to get reports about this now obviously purchasers were notified by Intel at all

I still don’t really understand that.

There’s still something not right about all of this.

 

[SpyderTracks]

GN have a response

This is really bad imho. They tagged on the oxidation statement as an additional lately

and then gave more in depth responses to the voltage correction BUT ONLY ON REDDIT

So there is no one source where they have a statement for all

 

[TonyCarter]

What a fustercluck!!!

 

[SpyderTracks]

For anyone reading this, I know I’m very aggressively anti Intel.

But if you’re still in warranty, and you’re having any kind of stability issues, open a thread on here or contact PCS, a clean install to be sure software is as clean as possible and then running some basic benchmark tests should tell you if there is an issue with your silicon. PCS have excellent RMA processes, you won’t be left in the dark

I would have thought if anyone is affected, even if your warranty with PCS has expired, I would run it past them first, think it’s important they can get a sense of numbers affected and I would have thought they’d have logs of serial numbers perhaps which may be of value to find if particular batches are affected more than others.

But even if your warranty has expired with PCS, you’ll be covered by parts warranty with Intel directly. I would fully expect by the time this new micro code is released, Intel will have caught up and put a dedicated page up to make it easier for people specifically related to this.

 

[SpyderTracks]

Jay coming at this from the SIs standpoint, worth a watch. I wouldn’t put PCS in the bracket where they’ll refer you to Intel, do contact PCS first

He also has suggestions on limits to apply now before the microcode update rolls out

 

[AccidentalDenz]

From the outside looking in, it seems as though one of the major failings of modern capitalism is biting Intel on the bum - extreme short termism where the next earnings call is all that matters. I don't follow the discourse around Intel all that closely, but I am aware they've barely innovated in 10+ years and have largely coasted by on the previously built reputation which allowed them to rake in the big bucks for that timeframe, but is now becoming a major issue.

Slightly different reasons, but the same basic idea of maximising short-term profits over longer term sustainability is causing issues for Boeing. They seemingly cut corners with design QA and airlines are turning to Airbus and other alternatives because the issues they've been having for a while are becoming more and more pronounced as it now seems barely a month goes by without an engire fire or a window blowing out mid-flight, etc.

 

[SpyderTracks]

These reports are hitting after Intels announcement they’re working on a mitigation, seems a lot of people aren’t convinced

Unreal Engine supervisor at ModelFarm blasts 50% failure rate with Intel chips — company switching to AMD's Ryzen 9 9950X, praises single-threaded performance

Criticism increases.

www.tomshardware.com

 

[HomerJ]

From the outside looking in, it seems as though one of the major failings of modern capitalism is biting Intel on the bum - extreme short termism where the next earnings call is all that matters. I don't follow the discourse around Intel all that closely, but I am aware they've barely innovated in 10+ years and have largely coasted by on the previously built reputation which allowed them to rake in the big bucks for that timeframe, but is now becoming a major issue.

Slightly different reasons, but the same basic idea of maximising short-term profits over longer term sustainability is causing issues for Boeing. They seemingly cut corners with design QA and airlines are turning to Airbus and other alternatives because the issues they've been having for a while are becoming more and more pronounced as it now seems barely a month goes by without an engire fire or a window blowing out mid-flight, etc.

it can also be said that game devs who made it the norm to release and then patch later has come to the cpu market as intel are being forced to patch an issue that could have been dealt with by having a bit more care instead of rushing.

 

[SpyderTracks]

it can also be said that devs who made it the norm to release and then patch later has come to the cpu market as intel are being forced to patch an issue that could have been dealt with by having a bit more care instead of rushing.

Although that would be a nice perspective, I don't think that's true of Intel.

Since 10th Gen it was clear they were pushing past acceptable limits. They were doing it to claim being competitive.

The entire point that they on the record told motherboard partners that unlimited power limits was "in spec" tells you all you need to know, there's absolutely no way that could ever be true of any silicon, they all have limits based on their binning and even the top most chip has limits they can't operate above.

They had to do this, they did so without any concern for their customers, that is evident based on how they've handled this situation.

The fact that they can't roll out the microcode update for over 2 weeks is because it isn't ready, and if they had truly identified the root cause as they say they have, there's no reason a tiny micro code update would take 2 weeks to write. And also if they did know the root cause as they state, then they would have announced exactly what voltage was being regulated, what new power limits were being put in place and how that affects boost clocks, because every single person or company with these chips is going to be worrying about those things.

They've purely stated this publicly due to the increasing pressure from big partners who are dumping their chips in their server farms.

I still don't believe this microcode update will stop the issue occurring, I expect it will just increase the time that degradation starts to occur on a brand new chip. And I still expect some chips to be faulty out of the box.

 

[SpyderTracks]

Great summary by Der8auer, skip to 10:44 related to the microcode update.

as he very correctly states, IF this microcode doe in fact stop degradation, it means that every cpu out there up to this point has had the degradation process started, it may well still be operating fine, but at a construct level, that degradation is fully underway albeit likely at a slowed pace.

 

[WilPad]

Im assuming PCS still do NOT recommend bios updating as a first step ?
if so , what do those of us with these Intel chips do for the best ?... how would the microcode get installed?

 

[SpyderTracks]

Im assuming PCS still do NOT recommend bios updating as a first step ?
if so , what do those of us with these Intel chips do for the best ?... how would the microcode get installed?

As always you need to contact PCS to get approval for any BIOS update to keep warranty.

 

[Scott]

Im assuming PCS still do NOT recommend bios updating as a first step ?
if so , what do those of us with these Intel chips do for the best ?... how would the microcode get installed?

It's not that it isn't recommended. It just needs gated and controlled. It's a very important part of your system and can go catastrophically wrong if not handled correctly. The warranty clause is in there for your own protection as well as PCS, contacting them opens up the channel of communication in order to assess if it's the best option for you and the method and version to go ahead with.

When the microcode update comes out, it will absolutely be recommended to do the upgrade..... but it should be done through the PCS channel for the above reasons, not to mention to keep your warranty intact.

 

[SpyderTracks]

The reality with any BIOS flash is that it can fully brick your motherboard so you can't boot.

Doing it with PCS approval immediately covers you if something should go wrong, so in the very worst case, if the board did brick (which is highly unlikely these days), PCS would just swap it out.

 

[WilPad]

Great ..thanks Guys... will wait for further info from PCS (thankfully my 13900K is running well and shows no signs of the degradation (YET ! ) )