Intel Core 13th and 14th Gen Desktop Instability Root Cause Update | Hacker News Reader

Intel Core 13th and 14th Gen Desktop Instability Root Cause Update

community.intel.com

48 pointstranspute17 comments

The fact they know about this hardware issue yet have refused to recall the faulty units has discouraged me from ever buying an Intel CPU again.

It's an expensive component that's meant to last 4-5 years at least.

There's a Gamers Nexus piece that goes into the tricking truth, stalling and memory holing of old press releases of Intel throughout this whole thing and it is actually quite ridiculous and galling. The 65watt thing people mentioned is just the latest party line, before those were definitely not susceptible according to Intel. I wouldn't surprised if this changed again at some point.

My read is they know all the chips are screwed and should do a recall but literally do not have enough stock for replacements and obviously don't want to take the hit. That's why they're giving people trying to RMA a runaround as well, they likely are prioritizing their big OEM customers over retail purchasers.

Intel says "all covered under warranty, everyone will be made whole", meanwhile people are being jerked around by service centers left and right. Last example DolphinChemist (top 1000 twitch streamer aka 0,01%er, not big not small) had Micro Center feed him "works fine in all our tests" line for couple weeks before finally relenting and replacing broken CPU.

Intel has extended the warranty for those units by two years.

AFAIK AMD had a similar issue with elevated voltage burning 7000 series CPUs and also didn't recall them. Is there something I'm missing?

> Is there something I'm missing?

The fact that AMD has responded in three days [0] with an update, and stopped the problem in its tracks?

Original article is on 20230424, "We're investigating" update is on 20230425, the firmware update and "yes we screwed, this the cause, here's the fix" update is in 20230427, and case is closed.

[0]: https://www.tomshardware.com/news/amd-ryzen-7000-burning-out...

All most recalls do is notify owners of the item that it needs to be brought in for servicing. Rarely do recalls actually mean a full replacement or refund for the item. Just look at the auto industry where you bring the car in and have the item replaced/repaired. For Intel, they are providing code that corrects the issue. Once the code has been updated, there's nothing to prevent a 4-5 years at least lifespan.

Maybe you'd be better assuaged if they also sent every confirmed purchaser of an affected CPU a $10 DoorDash coupon?

> Once the code has been updated, there's nothing to prevent a 4-5 years at least

That is not a solution. That is a mitigation.

What if you bought a car with 65mpg that has an undisclosed engineering flaw where the engine will explode every 15,000 miles. The "solution" proposed by the manufacturer is to bring the car in for "repair" that reduces it to 40mpg, but the car will last a normal lifespan.

Did you not lose a considerable amount of value on your purchase?

is that even remotely close to a realistic comparison? the update from Intel provides no information on a negative performance from what is advertised on the tin. they provided results showing it had negligible effects on performance.

wild anecdotes are fun to use on the internet, but provide nothing substantive to the discussion. but if you want to continue to memify everything, then have a nice day

On cars, a recall is for a specific component, which is replaced. I've seen recalls for an airbag inflator, tow bar, and shifter cable. That to me sounds the same as replacing a CPU, RAM, a PCIe card, or a peripheral on a computer, just more doable as DIY labor.

Tesla is infamous for OTA updates whether you wanted it or not which is very much inline with what this fix is.

I think they would have a time arguing against a warranty claim for a failure related to this issue though.

> Intel’s internal testing comparing 0x12B microcode to 0x125 microcode indicates performance impact is within run-to-run variation

Sneaky wording. Aren't they supposed to be testing the version before 0x125, considering it is the first one that may have potentially negative effect on performance?

> eTVB Microcode algorithm which was allowing Intel® Core™ 13th and 14th Gen i9 desktop processors to operate at higher performance states even at high temperatures. a. Mitigation: microcode 0x125 (June 2024) addresses eTVB algorithm issue.

emphasis mine.

Does anyone know if Intels NUCs were affected by this?

Any raptor lake arch chips using 65W or more. Some of the lower SKUs like 13500 use overclocked Alder lake chips (check the stepping)

Those tend to use mobile processors, right? Per the article

> Intel® reaffirms that both Intel® Core™ 13th and 14th Gen mobile processors and future client product families – including the codename Lunar Lake and Arrow Lake families - are unaffected by the Vmin Shift Instability issue.

I think NUCs usually have laptop class chips at 45w TDP or lower so afaik no.

I got fucked over by this with a gaming computer. Never buying intel again

Reply on news.ycombinator.com