News Intel continues search for source of Core i9 chip crashes — issues statement about recommended BIOS settings to board partners

[Admin]

Intel has provided guidelines for BIOS settings that it recommends its board partners apply to their motherboards by default. These settings are intended to improve 13th and 14th Gen Core i9 CPU stability, though the investigation is ongoing.

Intel continues search for source of Core i9 chip crashes — issues statement about recommended BIOS settings to board partners : Read more

 

[rluker5]

Those Intel settings seem completely reasonable.
I'm only using 4 of the 10 AFAIK, but I would rather follow them all than have instability.
Unlike Intel's failsafe voltage. I would find some other way before setting my CPU to that.

 

[CerianK]

Although possibly useful for troubleshooting (and thus short-term mitigation where effective), C-States 'Enabled' should always be optional (i.e., never required to be enabled for stable operation).

 

[NeatOman]

"Intel continues search for source of Core i9" ... insert Taylor Swift lyrics.. It's me, hi, I'm the problem, it's me 😅

 

[thisisaname]

Someone is pushing the silicon a bit to hard, not sure if that is Intel or the motherboard manufacturers. Or a combination of the two.

 

[Amdlova]

The bestway to solve a i9 13900k 13900ks 14900k 14900ks is a micro-code update do downgrade to a i7.

When some one say 90 - 99 degree celcius it's normal... It's normal to my Eletric kettle.

 

[CmdrShepard]

In other words, they have no clue why people's systems are unstable and press interpreted their private "advice" to mainboard manufacturers that got leaked as "Intel blaming the mainboard manufacturers" for clickbait.

I bet it will turn out that most of the instabilities are due to inadequate power supplies and cooling, mixed with not knowing how to configure BIOS properly.

 

[TerryLaze]

Intel has provided guidelines for BIOS settings that it recommends its board partners apply to their motherboards by default. These settings are intended to improve 13th and 14th Gen Core i9 CPU stability, though the investigation is ongoing.

Intel continues search for source of Core i9 chip crashes — issues statement about recommended BIOS settings to board partners : Read more

https://en.wikipedia.org/wiki/Volt-ampere
Watts volts and amperes go hand in hand, you can't have unlimited power if both volts and amperes are limited, it automatically limits the power according to the simple formula in the wiki page.
P= V * I (power in watts equals volts times amperes)

These recommended settings reveal that Intel is not convinced unlimited power limits are the primary issue. For now, it seems like Intel is more concerned about other parameters that apparently affect stability, including Current Excursion Protection and TVB Voltage Optimizations. One interesting tidbit is that Intel does provide an amperage limit on the ICCMax setting that's much slower than the maximum 512 amps limit motherboard makers can use.

 

[Nicholas Steel]

Although possibly useful for troubleshooting (and thus short-term mitigation where effective), C-States 'Enabled' should always be optional (i.e., never required to be enabled for stable operation).

Yes, C-States is merely a Power Saving feature and disabling it does not run the CPU out of spec. This is the same with AMD CPU's.

 

[Pierce2623]

As someone who’s had a 13700k and 13900ks. It’s pretty clear to me, that the 5.8-6.0 frequency range just requires tighter binning than they’re applying if they want good stability. Any 13th gen and up k chip is good at 5.5-5.6, but for the chips that boost higher than that, they need to apply tighter binning. Period.

 

[Sleepy_Hollowed]

I'm going to have to say, maybe, just maybe, they should've relayed this as a mandatory default for all board makers to be certified/resellers, but what do I know.

I remember when there used to be those for safety of the hardware and, apparently, safety of your place.

That being said, I also remember when sometimes video cards would crash because some vendors would not follow PCI bus specs as they should have.

 

[bit_user]

I bet it will turn out that most of the instabilities are due to inadequate power supplies and cooling, mixed with not knowing how to configure BIOS properly.

If true, that would be easy to disprove. Also, there's nothing new about those things and therefore no reason why they would be causing a flurry of issues only now.

The fact that Intel is taking the problem seriously is the main thing that tells me there's something to it.

 

[TheHerald]

In other words, they have no clue why people's systems are unstable and press interpreted their private "advice" to mainboard manufacturers that got leaked as "Intel blaming the mainboard manufacturers" for clickbait.

I bet it will turn out that most of the instabilities are due to inadequate power supplies and cooling, mixed with not knowing how to configure BIOS properly.

Inadequate cooling shouldn't cause crashing. The cpus regulate themselves and thermal throttle. I used a 14900k on a single tower air cooler, 0 problems with crashing.

 

[TerryLaze]

As someone who’s had a 13700k and 13900ks. It’s pretty clear to me, that the 5.8-6.0 frequency range just requires tighter binning than they’re applying if they want good stability. Any 13th gen and up k chip is good at 5.5-5.6, but for the chips that boost higher than that, they need to apply tighter binning. Period.

Are you talking about all core?! Because intel doesn't sell them as 5.8 all core....
The 13900k is 5.8 for 1-2 cores if and when temps and power allow it.
(turbo 3 )

 

[TerryLaze]

The fact that Intel is taking the problem seriously is the main thing that tells me there's something to it.

Yup, intel is taking this serious by calling mobo makers dumba**es.
Basically if you look at the graph intel posted, all they are doing is to tell them to not disable the limits (tjmax and icc max) that provide the proper limits for all of the boosts to work properly without overshooting.

 

[CmdrShepard]

Inadequate cooling shouldn't cause crashing. The cpus regulate themselves and thermal throttle. I used a 14900k on a single tower air cooler, 0 problems with crashing.

Cooling by itself? No.

However, poor cooling can aggravate already bad situation.

 

[CmdrShepard]

The best part of this mess?

For 13th and 14th gen CPUs you can download a datasheet here and navigate to chapter 13.0 to read electrical specifications and find absolute min and max values.

For Sapphire Rapids CPUs (which also have X SKUs which are unlocked for O/C)?

No such documents to be found.

 

[PCWarrior]

Here is a video explaining the cause of the issue and an easy fix to solve it.

View: https://www.youtube.com/watch?v=NWD5BgYDs9Y


When the cpu is hit with a fluctuating heavy workload and the cpu is trying to hit 6GHz (or more) on 1-2 cores there are cases where the motherboard’s VRM fails to respond correctly to the transient and shoves a very large voltage to those 1-2 cores. This in turn creates a heat transient that causes the crashing. So at the end of the day it technically is cooling related.

 

[TerryLaze]

When the cpu is hit with a fluctuating heavy workload and the cpu is trying to hit 6GHz (or more) on 1-2 cores there are cases where the motherboard’s VRM fails to respond correctly to the transient and shoves a very large voltage to those 1-2 cores. This in turn creates a heat transient that causes the crashing. So at the end of the day it technically is cooling related.

This is just his guess though, he also reduces all of the things (max watts, max temps, max volts) by reducing the max clocks so it still could be anything.

 

[jeremyj_83]

Yup, intel is taking this serious by calling mobo makers dumba**es.
Basically if you look at the graph intel posted, all they are doing is to tell them to not disable the limits (tjmax and icc max) that provide the proper limits for all of the boosts to work properly without overshooting.

Intel is only caring about the now to fix some bad press. Had the ACTUALLY cared about this before they would have made motherboard vendors enforce their recommended power setting by default. However, Intel hasn't wanted them to do this as it lowers performance and that lower performance means they lose benchmarks. Instead they turned a blind eye to this and viewed these settings as the true default.

 

[CmdrShepard]

Here is a video explaining the cause of the issue and an easy fix to solve it.

/offtopic ON

God how I hate those imbecilic faces they put on YT thumbnails.

There ought to be a law against it.

/offtopic OFF

As I said, cooling on its own probably can't cause it but there are multiple factors at play and cooling is certainly one of them.

Those CPUs can draw a LOT of current in a very brief period of time (think microseconds!) and no matter how good your cooling is, it can't react nearly as fast as the CPU can ramp up the heat.

 

[TerryLaze]

Intel is only caring about the now to fix some bad press. Had the ACTUALLY cared about this before they would have made motherboard vendors enforce their recommended power setting by default. However, Intel hasn't wanted them to do this as it lowers performance and that lower performance means they lose benchmarks. Instead they turned a blind eye to this and viewed these settings as the true default.

If that were anywhere near true then intel wouldn't suggest baseline settings now...
Intel would have searched for the highest possible settings that would prevent unreal shader compiling to crash and would tell mobo makers to not exceed those, that way they would still keep the highest scores.
Instead intel is telling everybody to use baseline settings.

Half a brain would have told you that, still everybody keeps parroting this because nobody has any capability of logical thought.

 

[jeremyj_83]

If that were anywhere near true then intel wouldn't suggest baseline settings now...
Intel would have searched for the highest possible settings that would prevent unreal shader compiling to crash and would tell mobo makers to not exceed those, that way they would still keep the highest scores.
Instead intel is telling everybody to use baseline settings.

Half a brain would have told you that, still everybody keeps parroting this because nobody has any capability of logical thought.

Did you not see where I said "Intel is only caring about it now to fix some bad press"?

Intel only cared about benchmark results. Therefore unlimited power duration was a good thing as it maximized their benchmark results.

 

[bit_user]

Intel would have searched for the highest possible settings that would prevent unreal shader compiling to crash and would tell mobo makers to not exceed those, that way they would still keep the highest scores.
Instead intel is telling everybody to use baseline settings.

This might still happen. I think Intel is currently in "damage control" mode, right now, and blaming motherboards for exceeding the recommended settings is an easy way to shift the blame.

Once they have a better handle on the actual problem, they could issue new guidance on how to safely improve performance over their baseline specifications.

Half a brain would have told you that, still everybody keeps parroting this because nobody has any capability of logical thought.

As I'm sure you know, ad hominem attacks violate forum rules. Please address the claim, rather than attacking the poster making it.

 

[TerryLaze]

This might still happen. I think Intel is currently in "damage control" mode, right now, and blaming motherboards for exceeding the recommended settings is an easy way to shift the blame.

Sending a loose guideline to mobo makers to implement whenever they want and however they want can't really be called damage control ,we have seen how real damage control looks with the exploding ryzen cpus, one agesa update after the other and new bioses left and right.
This is just intel saying to leave the safety limits in place.