News Intel issues official statement on Core K-series crashes: stick to Intel's official power profiles

[helper800]

Your assertion is correct. Other than this cock-up, it has worked out well.
What i assert is that the problem is motherboard manufacturers enabling insane power profiles without informing the users or asking for consent! You have yet to blame Intel for allowing this. If there is any smoking gun it's that.

This is what I was implying without directly saying it because I am poor with words. That is what I meant when I was saying Intel allowed this to happen, they have the power to prevent it and did not do so. Intel allowed partners to default their CPUs on the motherboards more aggressive/most aggressive profiles. We arrive at a similar conclusion but come from opposite perspectives. Intel has got to have some sort of default CPU behavior profile, and imo, they have to make the motherboard makers use it as the default stock profile on a system unless the user wants to change to a higher performance profile. I understand that K sku CPUs are essentially "made to overclock," but that should not come at the cost of stability on a default motherboard partners profile. I see fault in Intel for allowing such a problem to occur when it seems obvious such a problem would crop up if left up to the motherboard makers. If the partners start flipping the bird to Intel, Intel will go back to making motherboards again.

 

[bit_user]

That's not forcing anything, it's just a statement to clear up a technical term.

What's the May 31st deadline, then? That sure sounds like an ultimatum, to me!

 

[TerryLaze]

What's the May 31st deadline, then? That sure sounds like an ultimatum, to me!

Where was that in this statement/article?!
Quote, or stop bothering people.

 

[bit_user]

Where was that in this statement/article?!

Here:

https://www.tomshardware.com/pc-com...y-hopes-to-fix-issues-with-some-core-i9-chips
​

Right in the headline: Intel reportedly demands all board partners implement Intel Default (Baseline) Profile by May 31 — company hopes to fix issues with some Core i9 chips

Quote, or stop bothering people.

I'm pretty sure I'm better than most people at citing sources. Given that I took this from a headline (sorry, I thought it was from this article but turned out to be another), I didn't think I needed to cite the source.

Sorry to bother you with troublesome questions. LOL.

 

[TerryLaze]

Here:

​

https://www.tomshardware.com/pc-com...y-hopes-to-fix-issues-with-some-core-i9-chips​

​

Right in the headline: Intel reportedly demands all board partners implement Intel Default (Baseline) Profile by May 31 — company hopes to fix issues with some Core i9 chips

So where is the actual report?! The title saying reportedly demands demands an actual report to be there, there isn't.
In fact what that article says is:
"Beware: This information comes from sources other than Intel, so until we get an official confirmation from Intel directly, take this information with a healthy dose of salt."
But you are apparently allergic to salt and prefer to jump on any bandwagon you come across and just spout accusations against intel, just because you don't like them or something.

Also completely different article an much older article, so in your mind older info is more relevant than newer info?! Wouldn't surprise me with the stuff you let out sometimes.

 

[bit_user]

So where is the actual report?! The title saying reportedly demands demands an actual report to be there, there isn't.
In fact what that article says is:
"Beware: This information comes from sources other than Intel, so until we get an official confirmation from Intel directly, take this information with a healthy dose of salt."
But you are apparently allergic to salt and prefer to jump on any bandwagon you come across and just spout accusations against intel, just because you don't like them or something.

Given it's in the freaking headline, I think it was fair to ask the question. It's fine if you don't know the answer, but you don't need to attack me for asking.

If it's true, then it would indeed mean that these aren't merely suggestions or recommendations. Of course, I would expect communication between Intel and its board partners to be covered by NDA, so we'll probably never know for sure.

 

[helper800]

But you are apparently allergic to salt and prefer to jump on any bandwagon you come across and just spout accusations against intel, just because you don't like them or something.

Everyone has a bad take at least every once and awhile, but I dont think @bit_user is on one here. There is "a report" that this is forced, as I said. Because of NDA we will probably never know the truth of the matter, however, when someone makes "a report" and remains anonymous, but is trustworthy enough to write an article about and put it in the headline, I would assume it's probably factual. I could be very easily wrong on that one factual premise, but everything I said still follows whether Intel is forcing its partners to put out a default, more baseline power profile, or not.

I actually like Intel and their products very much. I recommend them all the time in the systems sections based on pricing. My first real system was a 3570k on a z77 motherboard and have fond memories of OCing it. Please don't assume we are bad faith anti-Intel folks, in fact, because we care enough to take time out of day to comment on such things should show that we care enough that we want them to succeed and improve.

 

[Silas Sanchez]

The first thing about the Intels that caught my eye is they draw alot more power than AMD while still producing the same or slightly better results in these benchmarks. The fact that all that extra power and heat isn't doing anything was to me an instant red flag and shows their product is being taken way out of it's intended operation and way into the inefficient region of the productivity curve. I assume they simply have a inferior process technology.
It's a basic law really, like when some lithium cells boast high discharge currents which look good on graphs, yet produce alot of heat. The extra heat means its working too hard and for longevity isn't a good thing.

Can't see why people go Intel when amd has the upgrade path, and amd is bringing out a new powerhouse soon. My amd cpu barley draws over 200watts at heavy loads while getting 90-95% or even more the scores of water cooled setups, my ddr5 ram speed is just stock 4800 and my system is eye watering fast. From what I can tell 3600 is actually good enough if you dont mind a handful of frames lost in games. For my own general computing I saw absolutely no diff in 4800 vs 6000. So keep it all stock and it stands a much better chance of reliability, then 2-4 years later you drop a new CPU in and you will surpass the overclockings and have a good laugh.

 

[bit_user]

I actually like Intel and their products very much.

I'm not anti-Intel, but I believe consumers are best served by keeping these companies honest.

Please don't assume we are bad faith anti-Intel folks,

I don't really care what someone assumes. If they're going to make allegations that I'm posting in bad faith, the burden of proof is on them. I was asked for a source and I cited it.

Now, presuming someone were posting in defense of Intel, I'm sure they would like to change the topic of discussion from Intel to its haters. Let's not fall into this trap and get defensive when we've done nothing wrong. The focus needs to stay on Intel and its partners.

 

[helper800]

I'm not anti-Intel, but I believe consumers are best served by keeping these companies honest.


I don't really care what someone assumes. If they're going to make allegations that I'm posting in bad faith, the burden of proof is on them. I was asked for a source and I cited it.

Now, presuming someone were posting in defense of Intel, I'm sure they would like to change the topic of discussion from Intel to its haters. Let's not fall into this trap and get defensive when we've done nothing wrong. The focus needs to stay on Intel and its partners.

I agree for argumentations' sake, but I was trying to go beyond the argument and reach a place of understanding irrespective of the argument, which may be naive, but we all have our faults. I am glad we have folks like you in the comments section keeping us honest and grounded. This time we just so happen to agree, as we mostly do, and others like @35below0 disagree this time around. He has a good faith argument from a different perspective, and I can at least respect that. @TerryLaze seems to believe that we just want to dunk on Intel because corporation bad, but clearly that is not our intent. Whether or not I can attribute that to good or bad faith argumentation can only be guessed.

 

[35below0]

You mean intel will have some bragging rights of having a good 10-15% overclocking headroom while the competition has zero.

Only half kidding here because people will see how much cooler the CPUs will run and how much less power they will draw and will go "hmmmm I can push that some more" .
I mean 20% less power and 12°C lower temps for 8% less performance, is not a small difference either.
https://videocardz.com/newz/intel-b...rformance-loss-compared-to-asus-auto-settings

No, i don't think so.

You are right 100% but it will not be seen that way.

This time we just so happen to agree, as we mostly do, and others like @35below0 disagree this time around.

I think you're right about the wording being a little bit difficult. In a way i didn't disagree with what you were arguing, but the verb "allow" can imply control, coercion, or passive acknowledgment.

Arguing in text only can be limiting.
Mostly i was trying to pin onto intel exactly that which they were guilty of instead of letting the hype train roll.
And even though i made that comment about throwing rocks, i didn't actually mean you. But until you quoted it, it didn't occur to me that i implied you were throwing rocks too.

Text.
Sometimes it's best to relax, think it over and say the simplest thing that is still on point and moves the argument forward. But, you know...

 

[helper800]

No, i don't think so.

You are right 100% but it will not be seen that way.

I think you're right about the wording being a little bit difficult. In a way i didn't disagree with what you were arguing, but the verb "allow" can imply control, coercion, or passive acknowledgment.

Arguing in text only can be limiting.
Mostly i was trying to pin onto intel exactly that which they were guilty of instead of letting the hype train roll.
And even though i made that comment about throwing rocks, i didn't actually mean you. But until you quoted it, it didn't occur to me that i implied you were throwing rocks too.

Text.
Sometimes it's best to relax, think it over and say the simplest thing that is still on point and moves the argument forward. But, you know...

I am not worried at all by the rock throwing comment. I took it in jest and assumed the best, that you were referring to more than just me. I edited my message shortly after you quoted it saying as much. In my responses to your rock throwing quote I wrote;

I am not throwing rocks, but facts. Joking aside, this is just how I see it, just my opinion.

This is because I take conversation seriously but try not to take implications as to my own doings personally. As you say above, its too easy in writing to make a claim and have it be interpreted in the most negative light or different in intent. I try to assume the best good faith interpretations of what people are saying first and clarify meaning and intent later if its a sticking point.

 

[logainofhades]

HUB has been quite adamant that Intel is the blame for this mess.

View: https://www.youtube.com/watch?v=OdF5erDRO-c

 

[Lucky_SLS]

I get Intel's desperation and unlocking the limits to have all the advantage it can get against AMD, but mobo vendors took it too far by having those profile enabled in lower end mobo with poor VRMs. I still feel that mobo manufacturers should be blamed for not having a proper QA process.

Sure, intel gave the go ahead for unlimited power, but mobo manufacturers abused that by marketing and offering the same profiles in all their models.

Case in point - the profiles were not named "Intel extreme" but after their own OC style.

 

[-Fran-]

HUB has been quite adamant that Intel is the blame for this mess.

View: https://www.youtube.com/watch?v=OdF5erDRO-c

Because they are to blame.

At least, unless super hardline blind fanbois, everyone should agree they're not blameLESS, right?

Their communication when asked about the power limits and motherboard testing their answer has always been the same "all in spec". That's it.

How can you not then conclude "ah, so they have looked at what the motherboard vendors have been doing and are ok with it". This nonesense that "but they can't control what they do" is pure, unadultered, bullshoes of the highest brown purity. It's borderline delusional.

Regards.

 

[35below0]

I get Intel's desperation and unlocking the limits to have all the advantage it can get against AMD, but mobo vendors took it too far by having those profile enabled in lower end mobo with poor VRMs. I still feel that mobo manufacturers should be blamed for not having a proper QA process.

Sure, intel gave the go ahead for unlimited power, but mobo manufacturers abused that by marketing and offering the same profiles in all their models.

Case in point - the profiles were not named "Intel extreme" but after their own OC style.

Not only that, but they are default profiles. Most PC builders didn't know that just assembling a normal, non-overclocking monster PC, and switching it on also by default set it to "crazy overclock monster mode".

 

[Moonstick2]

188W has never been any number in any intel manual concerning power, this alone tells us that these numbers are random numbers that gigabyte, or whoever, came up with.
It also has the extreme as 253W maximum which is also wrong and has been known well earlier to be 320W.

Several sites said their Intel sources were broadly confirming the content of that table when it arrived.

And Intel themselves said "These 'Intel Baseline Profile' BIOS settings appear to be based on power delivery guidance previously provided by Intel..."

 

[Moonstick2]

A lot of the back and forth here appears to be about who should be blamed for the original situation, even though Intel themselves still don't know the cause.

The point of the article though is that the motherboard manufacturers have issued profiles based on previous Intel guidance to try and fix the problem, and Intel are now telling them to make their power profiles more aggressive.

 

[TerryLaze]

Several sites said their Intel sources were broadly confirming the content of that table when it arrived.

And Intel themselves said "These 'Intel Baseline Profile' BIOS settings appear to be based on power delivery guidance previously provided by Intel..."

Intel would definitely know so they would not say appear to be.
Also they wouldn't say provided by intel.
If you have a link from intel where they say it please do link it.

A lot of the back and forth here appears to be about who should be blamed for the original situation, even though Intel themselves still don't know the cause.

Everybody including intel knows that the blame is mostly on reviewers because they are going out of their way to find the mobos with the highest default settings and then they go out of their way again to go into the bios and apply xmp but not change any other settings, and then they go out of their way again again to run power virus software, just to show the most clickbaity results they possibly can.
So of course every mobo maker is going to try to push their settings as high as possible because every mobo maker needs the exposure that comes from the reviews, there are tons of people that will just copy the hardware from a review or youtube video 1:1 for their build.

The point of the article though is that the motherboard manufacturers have issued profiles based on previous Intel guidance to try and fix the problem, and Intel are now telling them to make their power profiles more aggressive.

No the point of the article is to generate clicks.
The point of the intel statement is to let mobo makers know that they can use settings that are overclocking but they should not disable the settings that actually protect the CPU from degradation.
What intel is saying here is, it's ok to overclock but don't disable the 100 degree temp limit or the 400A current limit because those are actually important for the safety of the CPU.

 

[Moonstick2]

Intel would definitely know so they would not say appear to be.
Also they wouldn't say provided by intel.
If you have a link from intel where they say it please do link it.

Those words are literally in the Intel press statement quoted in the article at the head of this thread, and on pretty much every tech press website covering the issue.

 

[Moonstick2]

What intel is saying here is, it's ok to overclock but don't disable the 100 degree temp limit or the 400A current limit because those are actually important for the safety of the CPU.

Where's the link/evidence that they were exceeding 100°C or 400 A?

If the limits are critical to processor safety, why do Intel allow them to be disabled?

Current Excursion Protection for example, listed in Intel's new table as Enable. Originally you could only disable it in K processors, but earlier this year Intel released new microcode to allow it to be disabled in non-K processors. I don't see why they'd do that if it was unreasonable to disable it because of processor safety.

 

[TerryLaze]

Where's the link/evidence that they were exceeding 100°C or 400 A?

I said that intel in this statement tells mobo makers to stay within limits, whether they were exceeding them or not, Current excursion (and icc max) that you talk about yourself is the Amp limit protection, if you disable it the CPU won't protect itself from too high amps.
400A is even on the table as a number.
The thermal part is the tjoffset, tjmax is set at 100degrees and if you have an offset of say 10 it will got o 110 degrees.

If the limits are critical to processor safety, why do Intel allow them to be disabled?

Current Excursion Protection for example, listed in Intel's new table as Enable. Originally you could only disable it in K processors, but earlier this year Intel released new microcode to allow it to be disabled in non-K processors. I don't see why they'd do that if it was unreasonable to disable it because of processor safety.

Being able to do something and having the mobo maker forcefully enable it for you without you even knowing are two different things, that's why.

They sell fully unlocked CPUs THAT YOU PAID A GOOD CHUNCK OF MORE MONEY FOR to be fully unlocked so why should intel steal from you by not giving you what you paid for?!

 

[35below0]

Where's the link/evidence that they were exceeding 100°C or 400 A?

If the limits are critical to processor safety, why do Intel allow them to be disabled?

Current Excursion Protection for example, listed in Intel's new table as Enable. Originally you could only disable it in K processors, but earlier this year Intel released new microcode to allow it to be disabled in non-K processors. I don't see why they'd do that if it was unreasonable to disable it because of processor safety.

Even shlubb z790 motherboards will allow risky settings as long as the CPU in them is a K. My z790 UD will let me sidestep thermal protection with only a mention in passing that the setting requires a high end cooling solution.
In other words, let it run up to TJmax and let the CPU throttle itself. If the cooler is good, more performance will be eeked out of it.

The vanilla settings for the motherboard are "Optimization" and the optimization is to merrily run a 13600K at 5Ghz. The "Unleash" profile is grayed out 😎 presumably because it needs a i7/i9 and because it's quite literal.

We can argue about settings being open to users or what intel chooses to allow, but a bunch of these "optimizations" are veering dangerously towards craziness.


What we as customers expect is to have latest technology automatically adjust clocks and things so that we don't need to overclock or take risks with hardware. It can be done for us "intelligently".
Instead, what we got was a factory nuclear shockwave overclock with the words "optimization" and a smiley face tacked onto it.

Being able to do something and having the mobo maker forcefully enable it for you without you even knowing are two different things, that's why.

yes yes yes yes yes yes yes yes yes

Forcefully enabling sensible settings that the user is later free to overrule at their own peril is acceptable. But not setting insane profiles as defaults.

 

[Moonstick2]

The thermal part is the tjoffset, tjmax is set at 100degrees and if you have an offset of say 10 it will got o 110 degrees.

But there's no evidence that even happened. Even Intel's leaked memo didn't mention Tjmax.

 

[Moonstick2]

Being able to do something and having the mobo maker forcefully enable it for you without you even knowing are two different things, that's why.

They sell fully unlocked CPUs THAT YOU PAID A GOOD CHUNCK OF MORE MONEY FOR to be fully unlocked so why should intel steal from you by not giving you what you paid for?!

Non-K processors are locked processors, not unlocked.

Intel issued a microcode update to allow CEP to be disabled on otherwise locked processors.

The point is that on the one hand motherboard makers have been accused of using "insane" and "crazy" settings, but just a couple of months ago Intel were quite happy to make one of those safety settings optional even on a non-K.