Intel has a Pretty Big Problem..

[Tetras]

EDIT: Just got to watch the latest GN video - interestingly there is a suggestion there that Intel might have updated the microcode to try and solve one issue and ended up exposing this issue in the process :s which might mean there is no one fix for all these CPUs as it seems the mitigations for one issue are incompatible with the mitigations for another issue :s

I think what GN suggested about the microcode issues being an indication that Intel don't really know what they're doing is the biggest concern I have.

I don't remember microcode being a big issue in the past, or hardly even mentioned on tech sites, especially in relation to there being serious bugs that can damage your hardware.

I wonder if the ability to deliver them via Windows Update has led to microcode being considered far more lazily than it was before, like it is akin to any other kind of software that you can release half-baked, but I suppose the increased complexity of modern CPUs might be a factor.

 

[StriderX]

Like every other company who thinks they've won and gotten so monumentally ahead of the competition... it fatigued itself and got fat.

 

[Twinz]

Its so simple, Intel say its a 125 watt CPU, they print that on the box and on Intel.arc, for marketing purposes but what they are telling motherboard vendors is to default to 253 watt or higher.

Stop it, 125 watts period.

If the problem is single-core suicide-boost-voltages, that could still happen within a 125w envelope.

Since I am now more paranoid than I have ever been about voltage, I poked around in the BIOS of my Asus Hero and found the setting that limits max voltage requests. I set it to 1.4V because that was Buildzoid's *guess* as a safe limit.

It still boosts to the previous 5.5 limit I set a while ago, but now it doesn't hold it as consistently under full load.

It's annoying that we have to babysit these things now. I hope Intel can actually release a test that reliably identifies degraded CPUs.

 

[mrk]

As many current Intel users won't be changeing to AMD will they instead decide not to upgrade. This will depress CPU sales and obviously sales / profits of retailers. This mess has huge implications over a wide area.

Anyone on 12th gen systems likely won't need an upgrade for many more years so loads of time to see what's what.

Hell Qualcomm with their new chips for laptops and other devices is really compelling if they move into desktop space too.

I won't be upgrading my 12700KF build for many years now as just no need, I'm not CPU bound in games at 4K and none of the apps I use are bound by CPU performance either as everything is now GPU biased so as long as the baseline is excellent which any 12th gen or above is, then all's good really.

 

[Chud_Jenny]

Intel just might be F***ed - Tech News Aug 4

Intel just might be ****** - Tech News Aug 4▷ MY STORE - shirts, pint glasses & hoodies: http://paulshardware.net⇨ Sponsor: Thermaltake EX Fans | CT120 http...

www.youtube.com

Check out my new AMD Gaming rig!

I just cant trust Intel right now so I switched my workstation PC back to an AMD based system. Sponsored Links:New Users sign up for a Free World of Warships...

www.youtube.com

 

[Grim5]

A second law firm has started a class action lawsuit against Intel

Second Law Firm Begins Class Action Lawsuit Investigation On Intel On 14th & 13th Gen CPU Instability Issues

Intel's 14th & 13th Gen CPU instability issues are now facing a second class action lawsuit as another firm starts its investigation.

wccftech.com

 

[pastymuncher]

As many current Intel users won't be changeing to AMD will they instead decide not to upgrade. This will depress CPU sales and obviously sales / profits of retailers. This mess has huge implications over a wide area.

Personally I bought this 12600 non K as a stop gap with a view to upgrading to a newer release down the line hence the Z690 motherboard. When info on the 14th gen started appearing I intended to upgrade to the 14700k but once launched I just couldn't get past the power draw and temps these cpu's were throwing up so stuck with the 12600. I have been with Intel since Core 2 launched and grabbed top dog back from AMD. When the 3d versions of AMD's 9000 series launch arrives I will be going over to AMD and will not be going back to Intel for a long long time, if at all. Quite frankly Intel's actions disgust me. Refusing RMA's, refusing to admit there is a widespread problem with their cpu's and issuing a recall and basically blaming anyone but themselves. At this stage I can't see how anybody can defend them. They need to come out and admit responsibility and issue a recall although what they are going to replace the with I don't know, maybe a 12000 series of the same tier plus cash. Say you return a 13/14700k you get a 12700k back plus a cash amount as the 12000 series don't appear to be affected. They need to do the right thing by their customers!!

 

[Drumroll]

Intel just might be F***ed - Tech News Aug 4

Intel just might be ****** - Tech News Aug 4▷ MY STORE - shirts, pint glasses & hoodies: http://paulshardware.net⇨ Sponsor: Thermaltake EX Fans | CT120 http...

www.youtube.com

Check out my new AMD Gaming rig!

I just cant trust Intel right now so I switched my workstation PC back to an AMD based system. Sponsored Links:New Users sign up for a Free World of Warships...

www.youtube.com

And J2C is back to AMD .
Must’ve hurt.

 

[FreeStream]

Running a launch day 13500 here (although still unclear if this CPU is affected or not…) in a server, no issues thus far but I’ve also undervolted it by 50mV, so any voltage related degradation will be slowed.

Half tempted to watch for deals on a 14900, undervolt it, turn off TVB and limit the power to 125W.

That would I hope bring down the peak core voltages sufficiently so that any degradation doesn’t result in problems within a timeframe I care about..!

 

[DBT85]

I shall have to keep an eye on this with interest. My 14900k was planned to last as long as my 6700k had but I'm not sure it'll get there even with it running quite nicely at the moment with a 0.16 undervolt and otherwise default intel settings and hitting 39000 on R23.

Will spend some more time limiting the max voltage and toying with the ACLL.

Anyone know how to read the number on the IHS to find out manufacture dates or whatever.

 

[Jay85]

I shall have to keep an eye on this with interest. My 14900k was planned to last as long as my 6700k had but I'm not sure it'll get there even with it running quite nicely at the moment with a 0.16 undervolt and otherwise default intel settings and hitting 39000 on R23.

Will spend some more time limiting the max voltage and toying with the ACLL.

Anyone know how to read the number on the IHS to find out manufacture dates or whatever.

If you manually tweak your settings you're fine. It's the default stock settings set by the motherboard manu that is causing the issues.

Too much voltage and too much heat is basically what is happening. The latest bios update should help as well if you run these at stock.

From what i've read so far, it's the etvb that is one of the main culprits and also one of the previous bios fixes was actually injecting more voltage to try and stabilise the chips which in turn caused faster degradation.

 

[Rroff]

There is no guarantee manually imposing limits [entirely] prevents the issue, if this voltage / temperature problem is the cause it seems to be underlying unintended behaviour beyond just TVB.

 

[Blackjack Davy]

Like every other company who thinks they've won and gotten so monumentally ahead of the competition... it fatigued itself and got fat.

I'm waiting for the same thing to happen to Nvidia, hah.

 

[jigger]

There is no guarantee manually imposing limits [entirely] prevents the issue, if this voltage / temperature problem is the cause it seems to be underlying unintended behaviour beyond just TVB.

Intels power use is going to be a significant contributing factor.

 

[Tetras]

Running a launch day 13500 here (although still unclear if this CPU is affected or not…) in a server, no issues thus far but I’ve also undervolted it by 50mV, so any voltage related degradation will be slowed.

Pretty sure that the 13500 is not affected by either the overvolting, or the manufacturing fault. The 13500 is based on the alder lake 8+8 die (C0) and only one of these CPUs (13600) was included in GN's list of affected CPUs that they got from Intel, which I believe was just a mistake on Intel's part, where someone assumed the 13600 was a B0 die (since the 14600 is) and it is not.

 

[jigger]

Pretty sure that the 13500 is not affected by either the overvolting, or the manufacturing fault. The 13500 is based on the alder lake 8+8 die (C0) and only one of these CPUs (13600) was included in GN's list of affected CPUs that they got from Intel, which I believe was just a mistake on Intel's part, where someone assumed the 13600 was a B0 die (since the 14600 is) and it is not.

Aren’t some of the 13th and 14th actually rebranded 12th gen?

 

[Tetras]

Aren’t some of the 13th and 14th actually rebranded 12th gen?

Yeah, any i3 is, some of the i5-13400/14400 (including -F), all the i5-13500/i5-14500 and the i5-13600.

 

[wookiee87]

Aren’t some of the 13th and 14th actually rebranded 12th gen?

very early models of 13th were rebadged 12th gen's at heart, those 13th gen units may be un affected but given current events dont 100% bank on them being fine long term.

 

[jigger]

very early models of 13th were rebadged 12th gen's at heart, those 13th gen units may be un affected but given current events dont 100% bank on them being fine long term.

I wish my 12th gen would die TBH. Its an insult to the industry.

 

[wookiee87]

I wish my 12th gen would die TBH. Its an insult to the industry.

12th gen is all i'd recommend now!