Intel says their 13th/14th Gen instabilities are from 'elevated operating voltages stemming from a microcode algorithm'

[Bitter]

3 hours ago, leadeater said:

I don't think Intel is in as bad a situation as others may be thinking but I also don't think they are ready to anything about it yet and it seems it won't be next year either.

https://www.marketwatch.com/story/intel-to-suspend-dividend-cut-15-of-staff-upon-big-earnings-miss-d811a220

I mean that doesn't look great overall right now.

[leadeater]

30 minutes ago, Bitter said:

https://www.marketwatch.com/story/intel-to-suspend-dividend-cut-15-of-staff-upon-big-earnings-miss-d811a220

I mean that doesn't look great overall right now.

I'm well well but that still doesn't mean they are in terrible dire situation either though. For example there is a massive difference to Intel right now to AMD at their lowest, who still don't pay dividends at all and never have.

[Bitter]

1 hour ago, leadeater said:

I'm well well but that still doesn't mean they are in terrible dire situation either though. For example there is a massive difference to Intel right now to AMD at their lowest, who still don't pay dividends at all and never have.

I'm curious if they rebound next week or not. If they don't rebound then it's bad. If they do rebound then it was just a blip. I've seen this with plenty of other companies in similar situations where the rebound isn't a rebound but a slow steady year+ long march back to previous levels. I have a feeling that the chip problems and now other problems disclosed in their call are spooking investors away and the slide is going to hang for a while. Intel was for the longest time 'the gold standard' for corporate America. If your company bought PC's they had Intel processors and that made them a household name despite making a niche product that people don't really directly interact with. They're taking a hit right now and my opinion is that it's going to get worse when lawsuits start dropping or a FTC investigation fires up. If it can be proven they knew about their issues almost 2 years ago and failed to inform their investors correctly of possible large liabilities they were facing they could be in some hurt and be accused of misleading those shareholders which would really ruin their reputation among investors and have possible legal implications. They're walking a fine line.

[porina]

2 hours ago, Bitter said:

If it can be proven they knew about their issues almost 2 years ago and failed to inform their investors correctly of possible large liabilities they were facing they could be in some hurt and be accused of misleading those shareholders which would really ruin their reputation among investors and have possible legal implications.

There will be nuisance suits as that is routine business once you get to a certain size. We have two known potential areas of attack.

1, the oxidation issue, which they did know about a long time ago. It wasn't that big. Basically it was handled internally and that normally would have been the last of it. It only resurfaced due to the other issue. Basically if you somehow ended up with a defective CPU it would be covered by normal warranty. 

2, excess voltage degradation issue. From the outside it doesn't seem like they knew about this before given the long time needed to confirm it. Because it is a degradation over time, while the symptoms might be easily observable, it could be time consuming to prove the mechanism before giving a response. Now they have, and also are offering an extended warranty, what more could be realistically asked of Intel? While I certainly don't know all factors in law, a recall seems extremely unlikely as it is not safety critical. Maybe some will try for consequential loss, but that is way beyond my level to speculate much. If the microcode update materially changes the performance away from published specifications that could be problematic. Note I'm not including 3rd party benchmarks in that. Intel does not sell that, they sell their published spec.

 

It will be interesting to see what legal attack the law firm who is trying to set up a class action might use if they decide to proceed. 

[Mark Kaine]

Im wondering if this isnt the actually bigger scandal... 

 

 

Do we have a topic about this, when will GN, and LTT report about this, and will they still review intel products after this?

 

i mean, tldr: abusing "counterfeit laws" to get around legit warranty cases is surely a new low, even for intel?

 

(i don't think it is relevant they backpaddeled in that one case tbh) 

 

[Bitter]

I think GN did cover it?

[Mark Kaine]

34 minutes ago, Bitter said:

I think GN did cover it?

Maybe... i haven't really watched this 1 hour long (?) episode fully.  i just thought this would warrant its own cover story somehow... 

[dogwitch]

9 minutes ago, Mark Kaine said:

Maybe... i haven't really watched this 1 hour long (?) episode fully.  i just thought this would warrant its own cover story somehow... 

he now making jokes about how many times he edited his pin. that people called him out on that...

oh yeah he going to milk this hard.

[Quackers101]

while I think GN's cover of the situation gets too "personal" and clouded. the video from 1 day ago covers a lot of statements and correcting intel is doing, some that one would agree looks a bit shady for the consumer and going around the issue. but in general this will get worse with newer chips in future, with 2x or 3x the complexity added on from newer methods/tech that could cause a lot more issues forward. to other chips and brands that should also be hit on, or like with the SSD situations.

[dogwitch]

1 hour ago, Quackers101 said:

while I think GN's cover of the situation gets too "personal" and clouded. the video from 1 day ago covers a lot of statements and correcting intel is doing, some that one would agree looks a bit shady for the consumer and going around the issue. but in general this will get worse with newer chips in future, with 2x or 3x the complexity added on from newer methods/tech that could cause a lot more issues forward. to other chips and brands that should also be hit on, or like with the SSD situations.

i agree on personal part.

[Mark Kaine]

10 minutes ago, dogwitch said:

personal part.

he definitely doesn't like intel. but i also understand why, they're doing really shady stuff for a **long** time now...

 

 

the other issue ofc, as shady they might be, 2 manufacturers are definitely better than one for consumers... imagine we only had AMD... 

 

SAY HELLO TO $1000 CPUs!

[leadeater]

3 hours ago, Mark Kaine said:

SAY HELLO TO $1000 CPUs!

AMD started that 2 decades ago anyway 

[Dabombinable]

5 minutes ago, leadeater said:

AMD started that 2 decades ago anyway 

Just over 10 years ago as well, with the FX 9590 at launch...not sure how many people were suckered in by the "first 5GHz CPU" @ $920 (OEM).

[leadeater]

8 minutes ago, Dabombinable said:

Just over 10 years ago as well, with the FX 9590 at launch...not sure how many people were suckered in by the "first 5GHz CPU" @ $920 (OEM).

The AMD Athlon 64 FX-62 had MSRP of $1031 in 2006, I rounded up in decades  (18 years is close enough)

 

Edit:

Seems the FX-57 one year earlier was also $1031. AMD, kings of $$$ CPUs. Remembering stuff this old is hard lol.

[SpecTre12345]

2 minutes ago, leadeater said:

The AMD Athlon 64 FX-62 had MSRP of $1031 in 2006, I rounded up in decades  (18 years is close enough)

 

Intel Pentium Pro takes the cake though, back in 1995: 180MHz for $1075.

 

Not sure if that counts though because it was targeted for workstations and servers.

[leadeater]

56 minutes ago, BiG StroOnZ said:

 

Intel Pentium Pro takes the cake though, back in 1995: 180MHz for $1075.

 

Not sure if that counts though because it was targeted for workstations and servers.

I had a Pentium Pro... just not in 1995 haha. Was a great AoE2 computer but it was well old by the time I go it.

[Mark Kaine]

1 hour ago, Dabombinable said:

Just over 10 years ago as well, with the FX 9590 at launch...not sure how many people were suckered in by the "first 5GHz CPU" @ $920 (OEM).

10 years ago i was rocking a Athlon 64x2, dodged a bullet there!

[Dabombinable]

1 hour ago, Mark Kaine said:

10 years ago i was rocking a Athlon 64x2, dodged a bullet there!

I'd only fairly recently got my first modern desktop around then - the whole thing cost around $1200 (because in-store prices+built) but I could have gotten for around $900 online with deals. So many years on a Celeron D 2.4GHz and then Core Duo T2600.

[Mark Kaine]

7 hours ago, Dabombinable said:

I'd only fairly recently got my first modern desktop around then - the whole thing cost around $1200 (because in-store prices+built) but I could have gotten for around $900 online with deals. So many years on a Celeron D 2.4GHz and then Core Duo T2600.

i was honestly fine using this chip until 2017, it performed exactly like it did in 2007, did everything i wanted - then the asus board gave up the ghost... i really need to try to replace it someday, i miss Vista lol, and more importantly the softwares i was using (music editing programs)  

 

i also meanwhile bought a better Athlon for like 5 bucks

[Bitter]

6 hours ago, Mark Kaine said:

i was honestly fine using this chip until 2017, it performed exactly like it did in 2007, did everything i wanted - then the asus board gave up the ghost... i really need to try to replace it someday, i miss Vista lol, and more importantly the softwares i was using (music editing programs)  

 

i also meanwhile bought a better Athlon for like 5 bucks

For super basic stuff it's amazing how long hardware can go, basic website use, a little youtube, word processing, a Core2Quad or even a Core2Duo with some overclock still can do it but you definitely need a dGPU to handle some decoding for video playback.

[IseeFractals]

So lets recap the intel situation.

Stability issues for 13th/14th gen first started cropping up about 4 months ago,  and the problem has continued to get worse, and worse, and worse, all the while intel has tried to shift blame to motherboard partners for "unsafe voltages" But board partners have been playing the same voltage game since 9th or 10th gen, haven't they? And journalist, tech youtubers, and the community as a whole has spent YEARS talking about how insane the voltages, power consumption and temperatures of intel's chips have been....and every step of the way intel has been there to say "don't worry about it! That's within spec, all good fam!" 

So it's "all good" up until the point that shit starts failing, then "oh those pesky board partners, not abiding by the guidelines we didn't give them, or the corrections we didn't offer. Damn them!  DAMN THEM ALL TO HELL!" Even if voltage weren't being jacked up, it wouldn't "fix" the underlying problem, it would just slow it down. That....is grounds for a recall. At least it should be, but they refuse, as they've refused to honor RMA's, and even after they "identified" the issue for the 3rd or 4th time, they still thought it was reasonable for that fix to take weeks longer to release.....which means customers are left with the choice of either continuing to use a system that is actively degrading, or ya know....just let their very expensive hardware be a paperweight for a few weeks. 

12th/13th/14th gen are effectively the same chips. Their most recently released processors are actually a performance downgrade, which they obviously would have to be in trying to reframe development to some modicum of efficiency, which doesn't really bode well for future chips, does it? Seems kind of like a repeat of the 14nm++++++ days to me.

Battlemage, which is...how late? I don't even know anymore, is dead. It's not going to be competitive, but because Intel's delays on Arc were so stupidly long, this time around they're being held to their fab contract. Which means they're either going to have to release a product that no one will buy, which will pointlessly add more cost, or they'll just have to make some really cool keychains. 

Beast lake is canceled. What was supposed to be their "zen moment, their return to form" (for reals this time you guys!) just ain't happening, and from what i recall that's not the only hyped release that's been scrapped recently. 

AMD at it's largest is still only 1/12th the size of intel, and where AMD had design issues for 6 or 7 years, for intel it's been over a decade at this point, and they've only managed to stay competitive by raising the power levels, and sticking more and more cores onto their dies. Having 50% higher core count at double, triple (or more) the power consumption for single digit performance leads, in "some" workloads, isn't a sustainable model.  They've been operating on the assumption that they're too big to fail, and they've mostly been right.....but that's not going to work when they start screwing over customers AND board partners, particularly when when their product development seems to be getting worse. 

[BlueChinchillaEatingDorito]

And here we thought 12th gen was something they could build on... 

 

4 hours ago, IseeFractals said:

AMD at it's largest is still only 1/12th the size of intel, and where AMD had design issues for 6 or 7 years, for intel it's been over a decade at this point, and they've only managed to stay competitive by raising the power levels, and sticking more and more cores onto their dies. 

Ehh AMD's period of mediocrity was also over a decade. They thought throwing more cores at the problem with the entire Phenom, Phenom II, and FX line would close the gap but more cores don't help when each core was unimpressive. Even overclocking when gains were significant back then didn't tip the scale in their favour. Sort of like how we got 8 core chips on Android phones early on yet the experience was still as clunky. 

[porina]

6 hours ago, IseeFractals said:

AMD at it's largest is still only 1/12th the size of intel

How are you measuring this? One measure of company size is market cap, or what the stock market thinks a company is worth. Not saying it is a good measure, but it is one none the less. AMD are almost double Intel currently at 218B vs 115B. And this isn't only due to the recent troubles. AMD have been ahead of Intel on that measure for a long time. Then you get nvidia blowing both out of the water at 2.5T.

 

Other methods could be used, revenue, profits, head count?

[05032-Mendicant-Bias]

I feel people are overly pessimistic on Intel. AMD survived bulldozer, Intel will be fine.

 

After a decade of stock buybacks over R&D and fab investment, Intel fundamentals were in rough shape. Shareholders were allowed to plunder the company until it was no longer competitive. This recent coverup on 13 and 14 gen defect has hit the executive where it hurts, the stock price. I'm optimistic this will lead Intel to improve company culture, a slow and painful process.

 

Technology wise, as long as Intel keep pouring R&D money into CPUs, Fabs, GPUs and NPUs and related software and customer support, Intel is set to regain the ground lost to AMD and TSMC. With the advanced packaging, Intel is now ahead in that aspect. We'll see if Intel can introduce backplane power delivery and ribbonfet successfully.

 

[Vishera]

On 8/5/2024 at 8:35 AM, Dabombinable said:

Just over 10 years ago as well, with the FX 9590 at launch...not sure how many people were suckered in by the "first 5GHz CPU" @ $920 (OEM).

The thing is that the FX 9590 was just an overclocked FX 8300 and only 1% of the motherboards in the market had the VRMs to drive that power hog.

Anyway the single core performance was terrible and you would have been better off with a 2600K at a significantly lower price.