FAO Intel 13th/14th gen users

[Beany]

Yeah, this has been bubbling away under the surface in the tech press for a bit now, but it's starting to come out proper.

Intel has had manufacturing issue problem and OEM level (motherboard/microcode) config issues with it's 13th and 14th Gen processors (those using the Raptor Lake spin, which have a TDP of more than 65w - that we know of), causing crashes under certain common loads, and those processors that have had problems are likely permanently damaged.

I'd love to give more information - batch numbers, specific causes, resolutions - but for over a year, Intel themselves have been mealy-mouthed about it, putting out vague statements, post-facto updating them once the noise has died down etc. This video covers that. It's one of a series that will be upcoming with the same title - Scumbag Intel.

The only thing that seems consistent is that the Raptor Lake uArch seems to have been over-juiced to try to keep up in benchmarks with AMDs X3D parts (thanks to badly resting on their laurels before the Zen architecture was released,which was improved to be competitive and eventually better than Intels gear - boosting power was a panic response) and now, having sown that particular wind, they're reaping the whirlwind.

As the video title states, Intel have been, er, very scummy about this, and Steve Burke from GN, for want of a better phrase, is fucking done with it and is now going on the rampage.



For what it's worth, yes, AMD had problems last year with some of their processors getting overvolted and literally burning holes in themselves. The difference being that once it started occuring, AMD was engaged and up front.

They recognised that they hadn't been clear enough with power guidance to motherboard manufacturers (which they then clarified), and anyone who had the problem - which was a very small amount of devices from all accounts - were told that AMD had spoken to their motherboard partners and that they'd make it right (IE mobo and CPU replacements, no questions asked) and given there's been no blowback of people saying that hadn't happened, it seems they acted in good faith, went above and beyond, and fixed peoples problems.

And that's the difference - AMD held their hands up, said "fucking whoopsadoodle, send us your broken shit and we'll sort it" as soon as it was clear there was a real problem.

Intel has been obfuscating and denying and blaming others, and even now, they still haven't done the most basic of things, which is to reveal which batches were impacted by the via oxidation issue. And that's just one of the problems they're talking around and not actually answering questions about.

Those with a long enough memory will remember that Intel has a long, long history of monopolisation, including bribing major OEMs like Dell to not use AMD processors in their systems - so this behaviour isn't really a massive surprise.

"No-one got fired for buying IBM" comes to mind. Remember when everyone had an IBM on their desk? You probably don't, because they rested on their laurels and assumed everyone would just buy their stuff, while competitors were eating their lunch.

With AMD eating up datacentre and consumer space, and ARM suppliers now moving in on the Windows space, Intel are running the risk of going the same way.

[Barry]

Our latest batch of Dev PCs are using these I believe, I had a brief off the record chat with one of our in house techs. He looked nervous.

[Beany]

Our latest batch of Dev PCs are using these I believe, I had a brief off the record chat with one of our in house techs. He looked nervous.

Apparently the main thing is compress/decompress and similar single core, high load operations (where the CPU gets pumped with power to make the core boost hard as fuck), but the important part is to know who supplied the systems, and have their warranty service on speed dial.

If they were supplied by an ISV like Puget or Overclockers UK, then make a note of their support details and find out what they're doing about this.

If they were built in house, as above but for whomever supplied the CPUs, as if it's retail box - easy. If it was a 'tray' CPU (they'll know what this is), then trickier.

GN (as above) and Level1Techs are the main good sources for this, along with Buildzoid (Actually Hardcore Overclocking on Youtube) if they want to do some reading/listening on their general opinion on things. They should be the main sources for decently informed sources on this; ignore LTT, they just put out a video blaming motherboard manufacturers who, er, aren't the ones who let oxidized wafers out, or who wrote the microcode that requests the voltage from the CPU, possibly based on bad temperature data....

[Barry]

Yeah, I'm a leave it to the IT bods to worry about tbh, I'll just make sure I'm backed up/using cloud as much as reasonable

We're on UE5 so we are pushing things fairly hard, but the 4090s doing a lot of the lifting I imagine.

[Beany]

Ah, yeah, shame that one of the compress/decompress tools used in UE is a trigger for these things, allegedly.

https://www.radgametools.com/oodleintel.htm

Careful now

The info on RGTs page is a bit out of date, but it was one of the things that tipped people off to it.

[RobYob]

ignore LTT, they just put out a video blaming motherboard manufacturers who, er, aren't the ones who let oxidized wafers out, or who wrote the microcode that requests the voltage from the CPU, possibly based on bad temperature data....

Unduly harsh comment on LTT there Beany. Top comment on their video is from L1T's Wendel who collaborated with both GN and LTT on the Intel failure reports.

Anyway, just glad I slotted in a new Ryzen 5600 into my seven year old desktop a few weeks ago. Didn't even need my LTT screwdriver

[Beany]

The problem I have with LTTs video is that he - someone with big Intel sponsorship income - is the only person so far, out of the half dozen or so authorative sources working on this - who is putting the mobo makers front and centre as the cause of the problem, and only quietly acknowledging that it's Intel who effectively (and if this ever gets to court, I'd not be shocked if it's made explicit) told them that this was happy with the dangerously high power delivery. Intel claimed that this was all, in Intels words, 'in spec', even when they were blowing 300w and high voltages through the CPUs. It was "fine".

And it was (power usage and laughable inefficiency aside) - riiiight up until Intel got sloppy with their microcode and pushed out oxidised wafers into the supply chain. And left them in there for over a year....and then suddenly intel pushes out 'baseline specs' - and fails to clarify them properly, to still allow mobo makers to push limits; Asus and Gigabyte having different power levels for the 'baseline' spec is a perfect example of them looking like they're trying to solve the problem, but - and the cynic in me says 'on purpose' - not actually doing so, because that would impact benchmark results.

Techspot covers it pretty well here, including an extract from an Anandtech interview where Intel talk about what they consider to be 'in spec'
https://www.techspot.com/review/2836-in ... line-spec/

Ian Cutress: One of the things we've seen with the parts that we review is that we're taking consumer or workstation level motherboards from the likes of ASUS, ASRock, and such, and they are implementing their own values for that PL2 limit and also the turbo window – they might be pushing these values up until the maximum they can go, such as a (maximum) limit of 999 W for 4096 seconds. From your opinion, does this distort how we do reviews because it necessarily means that they are running out of Intel defined spec?

Guy Therien: Even with those values, you're not running out of spec, I want to make very clear – you're running in spec, but you are getting higher turbo duration.

We're going to be very crisp in our definition of what the difference between in-spec and out-of-spec is. There is an overclocking 'bit'/flag on our processors. Any change that requires you to set that overclocking bit to enable overclocking is considered out-of-spec operation. So if the motherboard manufacturer leaves a processor with its regular turbo values, but states that the power limit is 999W, that does not require a change in the overclocking bit, so it is in-spec."

Basically, if you've not changed the multiplier, then mobo manufacturers can, and will do, whatever they want, as according to Intel, it's fiiiiiine. The whole article is worth a read, as this isn't anything new or unknown - tech outlets have been complaining about Intels behaviour, and the guidance they provide to mobo makers, for knocking on a decade now. That chips have started failing is the inevitable endgame of this behaviour.

I'm quite happy being harsh on LTT as he's proven over the years to, lets say, not exactly be a bastion of integrity - so him pushing out a video that effectively deflects blame from Intel, exactly as Intel have already tried to do (it was the mobo makers who dunnit officer) would be exactly what I'd expect from both LTT and Intel - it's not like Intel don't have history of extremely illegal practises, which they have been very keen to defend.

https://en.wikipedia.org/wiki/Intel#Ant ... 80%932023)

This sort of behaviour is absolutely Intels bread and butter so extreme cynicism is entirely justified.

[Mito Man]

As a lay person in all this I’ve avoided Intel for the last 5 years. Just seem to make products seemingly no better than AMD but at a higher price so I’ve been an AMD person for a while now.
I think I’d like to move onto ARM, my Macbook M1, M2 have been great but the only problem is much of the software doesn’t run very well on OSX, or having to use Parallels negates any benefit. All the current Windows ARM based stuff seems to be touchscreen based Surface type shite.
When are they going to put their processors in a proper fat laptop?

[RobYob]

LTTs coverage in light of GNs absolutely teaming of Intel (seemingly absolutely right) and L1Ts coverage of a few weeks ago does seems very oddly kid gloved.

Not sure about the motivations, Intel haven't sponsored the "Tech upgrades" in a while, that's now AMD.
Their senior labs engineer is ex Intel GPU division.

Whatever the case GN appears to have evidence of prolonged systematic shitfuckery by Intel and good on them for investigating it.

[Beany]

GN are one of the major go-tos for people who are infuriated - they likely have sources within the teir 1 ODMs.

They are careful to consult legal before going on a tear, and going up against Intel is fucking big boy stuff, so you can bet they've done their due diligence. L1T is smaller and doesn't have the financial war chest GN does for hiring lawyers so they have to be a bit more conservative.

Anyway, guys....

https://x.com/datnofact/status/1820213413319962975


(in case it gets deleted, and to save duplication I've set the URL to x.com above)

Screenshot from 2024-08-05 13-12-03.png (75.84 KiB) Viewed 2173 times

[Explosive Newt]

So, hypothetically, if you had a 13th gen Intel CPU in your recently built home PC...

Do I just wait for some intel recall and replace it? Or continue using as per until it start to go weird?

[Simon]

First you need to complain at everyone on here who convinced you to buy Intel

[RobYob]

Do I just wait for some intel recall and replace it? Or continue using as per until it start to go weird?

I think until there is a recall the best you can do is keep an eye out for bios updates and make sure you're on the latest.

[Beany]

First you need to complain at everyone on here who convinced you to buy Intel

Don't look at me, I mentioned my cynicism on Ascenders thread, based on the info we had at the time

At the moment it's a case of wait and see. If you know how to change the performance settings in your BIOS then set them to something conservative, assuming you can find guidance on that, as the motherboard makers themselves don't seem to know what to do with Intels "recommended" settings.

If you don't know how to do that, get the latest BIOS versions (ones in mid august should contain mitigating patches) and keep an eye on the tech press to see if Intels position changes.

[Beany]

Speaking of updates, intel have clarified their incredibly fuzzy warranty extension that they edited repeatedly upon initial announcement:

https://community.intel.com/t5/Processo ... 853#M75727

Additional two years, and if we turned you down for warranty previously, we're so vewy vewy sowwy, please try again.

For users who are or have previously experienced instability symptoms on their Intel Core 13th/14th Gen Desktop processors and need to initiate the exchange process:

Boxed Processors – please contact Intel Customer Support for further assistance.
Tray Processors – please contact your place of purchase for further assistance.
OEM/System Integrator Intel Core 13th/14th Gen-powered desktop system – please contact your system manufacturer for further assistance.

If customers have experienced these instability symptoms on their 13th and/or 14th Gen desktop processors but were unsuccessful in prior RMAs we ask that they reach out to Intel Customer Support for further assistance and remediation.

So, still no batch numbers, etc, inviting speculation that all of the listed processors are impacted.

Well done Intel. Well done.