Intel processors crashing Unreal engine games (and others)

[Thunder 57]

I appreciate you are helping members that are unable to listen to the video.

I will be watching the whole thing including the sponsor ad because I support their efforts and consumer advocacy. The failure analysis lab is crazy expensive all by itself. GN donates to some cool charities too. I have bought a couple of their shirts: someday I'll get the modmat.

Besides no way am I missing out on - Thanks Steve, Back to you Steve.

These rumblings about the supply chain breaking down, make me think they must be dying faster than they can replace them. Do you think they changed the stress tests they use at Intel, and more are failing QA?

You're right about them donating. I recall GN asking the guy who wanted to gamble his paycheck about 14900K efficiency and Steve shut it down and asked him to donate his paycheck to Cat Angels (NC). They also put out written articles after their videos come out. I like this techtuber more than the rest.

 

[KompuKare]

Even of Intel fully refund a faulty CPU, with the high price of motherboards end users would still face quite a cost to swap platforms.

The lack of clear information means that any very recent buyers still within the return window should probably return both CPU and motherboard IMO.

 

[Kepler_L2]

What Intel didn’t write on Reddit but thinks internally - The search for the solution to the Raptor Lake S instabilities continues (Leak) | igor´sLAB

I already posted yesterday in a news item and briefly commented on what Intel officially published on the night of 22 July 2024 via Reddit as a kind of interim report. However…

www.igorslab.de

– Intel is validating a microcode update to limit VID requests above 1.55V as a potential future corrective action, targeted for production release in mid-August to NDA customers.
– Early testing by Intel on a small number of benchmarks indicates minimal performance impact due to this microcode change.
– While this microcode update addresses the elevated voltage aspect of this issue, further analysis is required to understand if this proposed mitigation addresses all scenarios.
– This microcode update, once validated and released, may not address existing systems in the field with instability symptoms.
– Systems which continue to exhibit symptoms associated with this issue should have the processor returned to Intel for RMA.

 

[poke01]

https://twitter.com/x/status/1815984426612056457

Why does it take so long to find the root cause?

-While Intel has confirmed elevated voltages impact the increase in Vmin, investigation continues in order to fully understand root cause and address other potential aspects of this issue.

 

[DAPUNISHER]

You're right about them donating. I recall GN asking the guy who wanted to gamble his paycheck about 14900K efficiency and Steve shut it down and asked him to donate his paycheck to a Cat Angels. They also put out written articles after their videos come out. I like this techtuber more than the rest.

CEO Snowball

Steve mentioned the used market issues this will create. I could not, in good conscience, sell a CPU I am not highly confident will serve the next owner well for years to come. My reputation and integrity mean far more to me than a few $100.

He is getting the quote for the FA lab, he is going to try to send off one for educational and edifications purposes. If any of you Anandtech ballers can afford to buy some merch or donate, now is the time.

The company that reached out to him is much bigger than MSI or Asus, I still have Dell on my bingo card.

Blanket do not buy 13or14 until further notice still in effect. He advised if someone bought the CPU recently to either do all the MC and bios stuff next month and put it out of their mind. Or return it. Anyone that has ruled out common instability issues should RMA pronto.

He threw a lot of shade at the dirtbag way they have gone about all of this. Along with all of the RMA rejection revisits ala Asus.

Epic going to AMD is another kick in the jimmy. Buying those high ASP SKUs too. The lost sales, the lost confidence, the avalanche of RMAs, the coming lawsuits, the nightmare media attention = category 5 crapstorm with massive financial damages that will take years to know the full extent.

More reports of 50% failure rates with significant performance stealing settings not resolving the problems.

Says this is Intel's biggest FUBAR in the 16yrs since he started GN.

@dmens told us Intel was falling apart. This must be him watching all of this go down

 

[KompuKare]

Could the big OEM be HP instead? After all Dell have been known as Intel's best friend which money could buy since the original Athlon days.

Anyway, that the comment about the large failure rate at a big SI at their intake QA: would love to know how they are testing.

No matter how good someone's intake QA is (and I appreciate they may have stepped it up just because of this) I cannot see it involving lengthy tests over days, so this System Integrator must have found some program or tool their intake QA uses.

Such a tool could really help end users.

About getting a full refund and considering a full platform swap: possibly going for a 12th gen replacement is the cheapest option for most customers. Normally resell value drops per "gen" (even counting rebrands), but in this case it might not.

 

[DAPUNISHER]

Jay just weighed in. As soon as I saw it I thought -this you Jay?

Now he is here to advocate for us and roast Intel? He needs to STFU and GTFO. Go to fat camp or something. His shilling and constantly sticking his finger in the wind antics are pathetic.

 

[igor_kavinski]

These rumblings about the supply chain breaking down, make me think they must be dying faster than they can replace them. Do you think they changed the stress tests they use at Intel, and more are failing QA?

I think Raptor Lake/RPL-R might be fundamentally broken and Intel probably has no real replacement for it. Not even Arrow Lake since that will be too limited in quantities due to the expensive process and packaging costs. Here's how I think things went down at Intel:

They realized with the oxidation issue in 13th gen that they need a newer stepping but possibly found out too late with thousands or even millions of 14th gen Refresh CPUs already made. So they decided not to create a bug free 13th gen K CPU stepping and just discontinued the entire 13th gen K line. But 14th gen is also hitting so close to its limits that it won't be able to last long unless it is downclocked/undervolted and delivers reduced performance. So they made a new plan. Enter the rumored Bartlett Lake. They probably started working on this as soon as they found out about the oxidation issue. They will refine their Intel 7 Ultra process even more, not skimp on validation this time and they will take out the E-cores so they only have to worry about the v/f power domain of P-cores and this would also reduce validation costs and time considerably. This helps them milk a couple more years of revenue out of existing LGA1700 mobos, both sold and unsold and also helps them deal with the supply issue for Arrow Lake CPUs which will need some time to ramp up enough in production so as not to need any contribution from LGA1700 CPU volumes.

By the way, those discontinued K CPUs that were already produced? I think they will show up as non-K 14th gen CPUs so these might turn out to have really good silicon compared to previous non-K CPUs, running in the best part of the v/f curve, delivering decent performance at acceptable power consumption. It's just a theory but if it's the case, this might turn out to be the silver lining in the Intel cloud of doom and dismay.

 

[Joe NYC]

Jay just weighed in. As soon as I saw it I thought -this you Jay?

Now he is here to advocate for us and roast Intel? He needs to STFU and GTFO. Go to fat camp or something. His shilling and constantly sticking his finger in the wind antics are pathetic.

Yup, exactly. Jay being JayTwoFaced with excessive shilling defeats the purpose of watching his videos for credible consumer advice.

 

[DAPUNISHER]

Could the big OEM be HP instead? After all Dell have been known as Intel's best friend which money could buy since the original Athlon days.

That's why I guessed Dell. They were also number one back then. Now they are number 3, and HP and Lenovo sell significantly more AMD based models. Who would be hurting worse than them, when it is 95% of their supply? We may find out through financials someday, who knows?

Anyway, that the comment about the large failure rate at a big SI at their intake QA: would love to know how they are testing.

No matter how good someone's intake QA is (and I appreciate they may have stepped it up just because of this) I cannot see it involving lengthy tests over days, so this System Integrator must have found some program or tool their intake QA uses.

Such a tool could really help end users.

That was the host's question on the full nerd podcast. His CPU passed all of the usual stress test everyone has relied on for years. But blackscreen with no log in fortnite, then EAC ban. He was wondering what new tools we will need too.

About getting a full refund and considering a full platform swap: possibly going for a 12th gen replacement is the cheapest option for most customers. Normally resell value drops per "gen" (even counting rebrands), but in this case it might not.

I agree, Alder Lake is fast and stable with DDR5 and games great. I can see it holding good value on the used market in a few years when no one that knows anything will give you more than $5 and a Snickers bar for raptor.

This has the potential to be the AMD graphics drivers problem, only with potentially far larger financial costs attached. What I mean is, it creates a stigma that follows the brand for years. Negatively impacting sales and resulting in users blaming anything that goes wrong, on your product/software even when it has nothing to do with the problems.

 

[Joe NYC]

These rumblings about the supply chain breaking down, make me think they must be dying faster than they can replace them. Do you think they changed the stress tests they use at Intel, and more are failing QA?

GN's first video had a statement I was baffled by. It said that a system integrator testing incoming (new) Intel CPUs was getting instability on (IIRC) 25% of them.

How is this possible? How is Intel sending the CPUs out like that? Untested or with very insufficient testing.

Edit: I see @KompuKare raised similar question above

Regarding "supply chain", (I don't remember the precise wording), but at the last Intel's earnings call, I had an impression that Intel sold a relatively large quantity of CPUs in Q1 from inventory, and that inventory was sold out in Q1. Going into Q2, Intel would have to fab all the Raptor Lake it intended to sell. Which now may be complicated by having to manufacture extra to fulfill current and potential RMAs.

 

[DAPUNISHER]

Yup, exactly. Jay being JayTwoFaced

Dang it Joe! Now we have his supervillain backstory waiting to unfold. He is going to go to fat camp, try to secretly make french fries in his room, have a grease fire, and burn half his face.

I don't watch his content anymore, and his crew are goofy goobers I would not trust to change a flat tire.

I doubt he added anything new to the conversation or has new information either. Just bandwagoning for dollars because he is a greedy goofy goober.

 

[igor_kavinski]

GN's first video had a statement I was baffled by. It said that a system integrator testing incoming (new) Intel CPUs was getting instability on (IIRC) 25% of them.

How is this possible? How is Intel sending the CPUs out like that? Untested or with very insufficient testing.

More importantly, why isn't the SI sharing their testing methodology with Intel so they can STOP shipping bad CPUs to begin with???

 

[Ranulf]

Jay just weighed in. As soon as I saw it I thought -this you Jay?

Now he is here to advocate for us and roast Intel? He needs to STFU and GTFO. Go to fat camp or something. His shilling and constantly sticking his finger in the wind antics are pathetic.

Heh, first comments I saw on that video from 8 months ago.. new comments from the last 2 weeks: "aged like milk" and "i guess you need to reverse back to ryzen".

 

[Kocicak]

Jay just weighed in. As soon as I saw it I thought -this you Jay?

Now he is here to advocate for us and roast Intel? He needs to STFU and GTFO. Go to fat camp or something. His shilling and constantly sticking his finger in the wind antics are pathetic.

You posted just his old video, you probably wanted to post the new one or both.

 

[Ranulf]

Tech Yes City has thought it was I/O issues for a year now with the newer Intel chips.

Recent video:

Last years video (he has a part two with more testing too):

 

[Joe NYC]

More importantly, why isn't the SI sharing their testing methodology with Intel so they can STOP shipping bad CPUs to begin with???

Intel may be trying to guard this tool, so that it does not get leaked to the public, which would allow customers to test their own CPUs, and possibly open the floodgates of RMAs, all confirmed by Intel's own tool.

 

[Timorous]

So pure speculation.

It seems all roads lead to rome (or in this case too much voltage). Microcode, motherboards, oxidation etc all seem to lead to more volts than the si can handle which causes premature degradation.

My speculation on the root cause of this is Intels speed bins being too lax. They are letting ropey si bin as an i9 when more stringent binning would kick it down to i7 or i5 bins. This means when that ropey si does get too many volts from whatever cause it degrades very rapidly and we get instability as a result.

So yea. I think Intel flew too close to the sun on their speed bins and even in a perfect world their lower quality bins would only just make their expected life span, throw in a few sources that increase voltages and all of a sudden si that may have just survived at stock with no more than spec settings now degrades rapidly with just marginally above spec voltages.

 

[Nothingness]

https://twitter.com/x/status/1815984426612056457

Why does it take so long to find the root cause?

-While Intel has confirmed elevated voltages impact the increase in Vmin, investigation continues in order to fully understand root cause and address other potential aspects of this issue.

Debugging SW can be incredibly difficult (which is why when I'm asked how long it will take to fix something, I always say from 2 minutes to 1 month). Hardware is even more difficult, so I wouldn't blame Intel for that (OTOH I hope they are not trying to hide anything, that could be devastating if it goes out the door).

 

[coercitiv]

Tech Yes City has thought it was I/O issues for a year now with the newer Intel chips.

Even in his latest video he's unable to explain what he was told in a coherent manner (timestamped video):

He says the following:

• I/O Hub was removed directly off the die
• it's not removed from the CPU itself, it's still on that silicon piece that is on the CPU PCB
• it's not like on the 10900K, where the I/O hub had a direct connection to the CPU

I hope you can see the huge contradiction here, something was removed from the die but is still on the die, and lacks a direct connection to the CPU while being part of the CPU.

 

[marees]

https://twitter.com/x/status/1815984426612056457

Why does it take so long to find the root cause?

-While Intel has confirmed elevated voltages impact the increase in Vmin, investigation continues in order to fully understand root cause and address other potential aspects of this issue.

- Intel observes a significant increase to the minimum operating voltage (Vmin) across multiple cores on returned affected processors from customers.

Is there a way that end-consumers can measure the increased min-voltage by themselves and check if their CPU has started degrading ?

 

[CakeMonster]

TYC rambles and has had some wild theories before that made me not at all trust him on deep technical issues. He does great stuff about deals, retro builds, restoration etc, but I would steer away from him when he's rambling about politics or the more technical stuff.

 

[blckgrffn]

Just because we created this satire urn design years ago and now its even more pertinent

 

[marcUK2]

That no major review site has attempted to benchmark the new baseline is ridiculous.

I know baseline is a moving target due Intel being evasive, but so much for independent reviewers!

Spent a lot of time looking for reviews after the supposed fix came out. Quite disgraceful - is everyone in intels pocket? The best i could find were reviewers applyinf the fix then tweaking the setting to run permanently at 253W...which is just intentionally misleading when its supposed to rapidly drop to 125W (in my case). Big difference between 35000CB and 22000CB when done accurately.

Fortunately because i bought a non-k processor and had a junky cooler to begin with I did some good setup before use and actually had a limit of 300A set manually from the start, and when I upgraded the cooler did it all again and set 325A manually as the 65W enforced setting is just ridiculous bad. 17000CB if im lucky. So i think i dodged the bullet on overvoltage for my CPU, but the default settings for my MB are basically unlocked Amps and Watts and it throttles in seconds.

 

[DAPUNISHER]

I put on the last of my warpaint when it was revealed they knew about the fab defects and were still denying RMAs for DC by the 1000s. It turns out Hardware Times never got the promised refund either.

This thread has obviously deviated from its topic, courtesy of the OP, who is clearly more interested in ragebait.

I would suggest the forum overseers shut this thread down.

There's nothing more to be discussed here.

ASUS has outed new BIOS with updated settings profile called Intel Baseline Profile.

View attachment 97318Finally this crap is going to end. Now Intel just needs to enforce this profile on initial setup across all vendors.