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

[hotaru251]

Then why didn't Intel do it at launch ?

intel doesn't try to control bios unlike nvidias iron grip.

only time intel really steps in is if its harming them. (instability can lsoe customers to team red and afaik only other time was when intel banned em from letting MB makers OC non k cpus as that impacted their sales)

amd & intel on cpu side generally let MB's do w/e as it rarely causes issues. (amd similarly stepped in when MB makers were letting ppl OC x3d cpus and damaging them)

As I said multiple times if the CPU is unlocked then there's no such thing as "out of specs".

...yes there is.

If you buy an (hypothetical) unlocked car are you going to run it at 500+mph? no. you'd likely have problems if you tried.

same for cpu's. unlocked means you can change all its settigns...but it doesnt guarantee anything over the specs out of the box. (and w/ silicon lottery some cpu can barely oc at all)

Wouldn't the motherboards be made and the bios be set prior to reviewers receiving them?

reviewers generally get pre-launch bios so they can have their stuff out on time but the launch bios may get changed after that.

again a reviewer who cares about reviewing will make sure to set the bios to make sure nothing weird is used (as not a fair thing if 1 mb allows you to go unchecked while another keeps it tame by default)

Motherboard manufacturers exceeded it some 20x

iirc a yr or so ago didn't asus have a moment in spotlight when they had some high as heck stock voltage on amd boards where you would have soemthing like .9v in bios but it would actually be 1.2v or so? (again not only brand that has higher than settigns say just that its not a new issue for MB's to be out of spec)

Not if it happens with the default profile.

you honestly think Intel checks every mb's default settings?

For sure is not user's fault but surely it is Intel that leaved mb manufacturers to do so, Intel that not restricted power, current, voltages and do not imposed safe defaults

....so you are asking Intel to become Nvidia and stifle the industry???

Nvidias destroyed the AIB side of GPU because they are overbearing. NOBODY wants that on CPU side.

All to cheat users using reviewers.

assumptions are bad. also Intel cheating at times isnt unheard of (recall the "chilled" cpu that hid what was cooling it from few yrs ago)
however you assume all reviewers would risk their credibility for intel?

Lets see if, with new BIOSes, the Extreme profile will be the default profile.
Imho what must be reviewed should be the default profile, without changing before the review.

depends on if you are reviewing the cpu or the mb.

its a well known thing out of the box default mb settings use high voltage.
also nobody in right mind would ever have an "extreme" setting as default.

It is Intel that allow this.

you seem to not undersatnd how intel and MB's relationship works...
they are NOT nvidia. They allow some freedoms so MB's can have differences between them.
If Intel did control them to extent you want you'd have much higher cost MB's as there would be no point to having many sku as you are limited (this is why GPU's are barely profitable for aib anymore)

Intel give mobo makers the permission to do this

you ignore point that Intel LITERALLY says if you OC your cpu thats not their fault you took the risk.

Intel notified users of the risk and legally are in the clear.
Now MB vendors? they do NOT warn you it's OC'd by default. That would be the guilty party...not Intel.

"but but intel allows it"
yes, because they have some trust in their partners to be reasonable.

I think that it is the motherboards that are failing to deliver the voltage under heavy load due to their relatively weaker power delivery

for sure MB can "fail to provide" in their default bios settings.

When I built my 5700x system the MSI x570s-edge wifi MB's default bios caused crashes (eventually pinpointed thank you windows crash logging) due to it having too low a voltage for the RAM settings (so when you ended up using enoguh ram system would crash) manually adjusted that and ran fine since.

MB vendors just need to agree on "the specs of the cpu maker as default" and maybe have a flash screen on 1st boot that notifies that you can enter bios and change profile if you want a bit better performance.

 

[Eximo]

Something interesting to note related to this:

The first diy PC I built had a 4770k and a Z87 Deluxe, which had 16 phase power delivery.
My current 13900kf is on an Asus Prime Z690P that has 14+1 phase power delivery, yet uses like 4x the power under heavy loads and Asus "default" settings.

I think that it is the motherboards that are failing to deliver the voltage under heavy load due to their relatively weaker power delivery. So it could help to push those undersized power delivery systems harder with an LLC that allows less vdroop.

Not sure comparing VRMs from such wildly different power requirements makes much sense.

Each of those phases was likely a much weaker VRM, so a direct comparison is difficult without knowing the specs of the VRM chips themselves.

 

[rluker5]

Not sure comparing VRMs from such wildly different power requirements makes much sense.

Each of those phases was likely a much weaker VRM, so a direct comparison is difficult without knowing the specs of the VRM chips themselves.

You mean like a 60-90 amp per phase Digi + VRM vs a 60-105 amp per phase DrMos?
They aren't that different, not 3x-4x different per phase.

Once again, lower powered chips are not having problems and some higher powered ones having problems at high consumption with high vdroop settings.
The difference seems to be power. And, no, power delivery has not changed as much as power consumption.

 

[helper800]

You mean like a 60-90 amp per phase Digi + VRM vs a 60-105 amp per phase DrMos?
They aren't that different, not 3x-4x different per phase.

Once again, lower powered chips are not having problems and some higher powered ones having problems at high consumption with high vdroop settings.
The difference seems to be power. And, no, power delivery has not changed as much as power consumption.

Your ASUS Z87 Deluxe motherboard has 60 amp VRMs with 16 phases. Most motherboards today have around the same amount of phases, but have 75-105 amp VRMs per phase. They also have much more advanced chips controlling all that power delivery. There are also many more components that go into a motherboards power delivery

 

[NinoPino]

you honestly think Intel checks every mb's default settings?

I have not said that, I wrote that for me I isn't the user to blame because it is normal for the user to expect that in the default profile there is no overclock.

....so you are asking Intel to become Nvidia and stifle the industry???

Nvidias destroyed the AIB side of GPU because they are overbearing. NOBODY wants that on CPU side.

I expect that Intel supervises the actions of his partners and if see actions that can damage the users or Intel's image than act to remedy it. If really are years that motherboard makers do whatever they want, there's no way Intel doesn't know and consequentially is Intel to blame.

assumptions are bad. also Intel cheating at times isnt unheard of (recall the "chilled" cpu that hid what was cooling it from few yrs ago)
however you assume all reviewers would risk their credibility for intel?

It is what just happened, a lot of reviewers consider the 13900 and 14900 the best CPUs also with the power consumption of a oven and an advantage on the competitor of few percentage points.
The only looser here is the user.

depends on if you are reviewing the cpu or the mb.

its a well known thing out of the box default mb settings use high voltage.
also nobody in right mind would ever have an "extreme" setting as default.

Of course I am talking of CPU reviews.
I doubt that motherboards use overvoltages on default as you wrote.

you seem to not undersatnd how intel and MB's relationship works...
they are NOT nvidia. They allow some freedoms so MB's can have differences between them.

motherboards can differentiate on a lot of characteristics, not only overclock capabilities. There is no problem if Intel impose some rules on the defaults.
I'm not saying to prohibit overclock but on defaults must be safe settings.

If Intel did control them to extent you want you'd have much higher cost MB's as there would be no point to having many sku as you are limited (this is why GPU's are barely profitable for aib anymore)

How imposing safe defaults can lead to higher costs ?
Graphics cards and motherboards are completely different and profitability cannot be compared.

you ignore point that Intel LITERALLY says if you OC your cpu thats not their fault you took the risk.

I do not ignore this, I agree with Intel on this.

Intel notified users of the risk and legally are in the clear.

ok.

Now MB vendors? they do NOT warn you it's OC'd by default. That would be the guilty party...not Intel.

"but but intel allows it"
yes, because they have some trust in their partners to be reasonable.

Not in my opinion, Intel not only permitted but also encouraged this behaviour to obtain an advantage on reviews.

MB vendors just need to agree on "the specs of the cpu maker as default"...

I agree.

 

[CmdrShepard]

I don't care what you care about.

That's because you are arguing from the position of ignorance.

Intel sold CPUs unlocked for overclocking. They did not oveclock them into instability.

As I said, you have no idea what you are talking about.

Exhibit A:

https://imgur.com/a/ASzJyAb

View: https://imgur.com/a/ASzJyAb


I set PL1 and PL2 limits for my Xeon w5-2455X to 300 W and 360 W in BIOS (up from defaults of 200W and 240W). So why are they both set to 500 W now you ask?

In XTU there is a button front center (in simple view, not advanced view mind you!) which says "Optimize Power & Current Limits". Out of curiosity I clicked it, and you can see the result on the image.

Yes, Intel XTU "optimized" my PL1 and PL2 to 500 W!

So please stop with the BS about Intel not pushing unsafe values themselves.

Intel could hardly sell unlocked CPUs that don't allow overclocking? Or only allow half overclocking?
People who bought those CPUs may have overclocked them, but even those who didn't suffered instability.

Or you know, they could sell tighly binned chips at a higher price with higher overclocking limits and potential which are UNLOCKED BUT NOT UNLIMITED.

Exhibit B:

https://imgur.com/a/lrxiMfU

View: https://imgur.com/a/lrxiMfU


Can I really run this CPU at 16 GHz SAFELY?!?

If the "optimized" power is 500 W for meager 4 GHz would they set 2,000 W for 16 GHz?

It's a rhetorical question, hope it proves the point which was -- if Intel wanted safe limits they would limit what values can be set in the CPU itself to begin with. Those values could've still been considerably higher for K SKUs compared to non-K SKUs, BUT STILL WITHIN SPEC.

Motherboard manufacturers exceeded it some 20x

Then I guess I am lucky that in the case of my CPU Intel XTU has exceeded spec by only 2.5x, right? 🙄

Intel i7s and i9s will perform well in benchmarks regardless. There are very few scenarios that gimp those CPUs.

You sound more and more like an Intel fanboy defending them. Multi-threading scenario loses 20% of performance. Most loads nowadays are multi-threaded so everything is affected.

This overboosting is done without informing the user or asking the user to confirm they want their system pushed over a cliff in order to fly faster.

Again you have no idea what you are talking about. When you install a CPU on a mainboard for the first time you are asked to enter BIOS to configure it. If you ignore that it's on you.

Intel can only be blamed for tolerating this situation for so long. But in their defence, they don't want to antagonize motherboard vendors. And CPUs weren't crashing until very recently so...

No, Intel can be blamed for allowing setting of unsafe values which are out of spec even for unlocked CPUs and for their XTU which also "optimizes" stuff to a crisp if allowed.

 

[35below0]

well, letting the default going (way) higher than intel extreme profile is surely a mishap from Motherboard vendors, BUT ! Now in order to fix the madness, Intel didn't ask vendors to run at the norminal "performance" or "extreme" profile which are both in spec, but ask them to go for a "baseline" / "fail safe" profile which actually stops the CPU from going full advertised frequencies for a lot of case, is Intel's fault

This is a good point.
Right now, the damage is done. The CPUs need to be stabilized. Benchmark results have been set and most Raptors sold. A new chipset and CPU gen is coming, so in many ways both Intel and the others will move on.

The good news is even though i9s will take a performance hit, they still have enough. A lot.

We don't know how much of advertized performace was unattainable without overdriving CPUs. Probably Intel never exceeded their own profiles (and why would they), but they knew that one of the marketing points was overclocking potential. Here they stated their usual "You overclock at your own risk", but did they know motherboard vendors would be going wild? Yes.

The reason Intel did not step in to reign in Asus, Gigabyte and MSi is because i9 and i7 chips weren't crashing. Once that started happening and was confirmed, AND Intel was blamed, they did step in.
Now they want a safe profile to be set as defaut, rather than one of the other Intel profiles.
Those can still be set by the user, or exceeded by the user.

Not if it happens with the default profile.

The default profile was not set by Intel.
The default profile was not one of the 4 Intel profiles.

I have not said that, I wrote that for me I isn't the user to blame because it is normal for the user to expect that in the default profile there is no overclock.

I agree.
Intel did not overclock the CPUs to hell, Gigabyte did. Why doesn't Intel stop them? Intel can't. How can they?

The K is unlocked. It's not unlocked for users but locked for motherboard manufacturers, it's unlocked for everyone.
It would not be better if motherboard manufacturers were not allowed to control the CPU clocks. There is no scenario in which that is desirable.
Think.

Your ASUS Z87 Deluxe motherboard has 60 amp VRMs with 16 phases. Most motherboards today have around the same amount of phases, but have 75-105 amp VRMs per phase. They also have much more advanced chips controlling all that power delivery. There are also many more components that go into a motherboards power delivery

Fair to say that about the more advanced components, but they still cheap out on a lot of stuff.
Mainly though, far too many motherboard ship with 60A or even 55A FETs. If only 75A were standard, but it isn't.
Only highest end motherboards will have 105A.

I expect that Intel supervises the actions of his partners and if see actions that can damage the users or Intel's image than act to remedy it.

They do.
Intel designs the CPUs and chipsets and publishes specs. One cannot make motherboards without sticking to Intel specs.

But without freedom to improvise, all Gigabyte, MSi, and others could do is manufacture one motherboard. Same exact one.
Again, think.

It is what just happened, a lot of reviewers consider the 13900 and 14900 the best CPUs also with the power consumption of a oven and an advantage on the competitor of few percentage points.
The only looser here is the user.

They are the best CPUs, the most efficient CPUs and least power hungry CPUs. Just because the i9 can take 350w doesn't mean it has to.
What happens is the CPU is fed unlimited power until it reaches TJMAX and throttles automatically. This way power consumption depends on the cooler. Better cooling can allow more power to be used.

This does not change efficiency. To gain higher clocks, more and more power needs to be wasted with diminishing returns.
This is why intel does not try to claim their i9 breaks the 6Ghz barrier.

They would love to brag about this, but it can only be attained for a few minutes, wastes huge power, and damages the chip. An overclocker will do this, but Intel would rather not.

If you want absolute max performance, you must use the power of an oven.
But the i9 can do the same work an i5 could, and use less power doing so. That is the i9 advantage, but people focus on max performance. Why? Because it's the most powerful consumer CPU and they want to know where the limits are, and break them.

Nobody who bought an i9 is a loser.
Except for this instability and crashes problem.
That was caused by Asus and Gigabyte ignoring 4 Intel profiles and choosing their own as default.
The unlimited profile was set as default. That's the problem.

I'm not saying to prohibit overclock but on defaults must be safe settings.

Yes and it wasn't. Intel spec allows this, and Raptor CPUs can withstand a lot of abuse. Motherboard manufacturers went too far. That became the problem so Intel stepped in to stop it.
They could have stopped it a year or two ago, but PCs weren't crashing so nobody cared.

Not in my opinion, Intel not only permitted but also encouraged this behaviour to obtain an advantage on reviews.

Show proof.
Intel spec does permit unlimited profiles, but prove they encouraged this and prove it gave Intel an advantage in reviews.

Yes, Intel XTU "optimized" my PL1 and PL2 to 500 W!

You used Intel's Overclocking tool to overclock your CPU. Why is this what you think it is? You want the OC tool to only OC to safe defaults? That's fair. I can agree to that.

So please stop with the BS about Intel not pushing unsafe values themselves.

You are flat out lying here.

Intel has been accused of "pushing unsafe values" by allowing motherboard manufacturers to set wild defaults and not doing enough, or anything to prevent them from using wild unlimited profiles as default.

As i wrote in the rest of this post, Intel spec allows for unlimited profiles. Same Intel spec offers 4 profiles that can be set as default on a motherboard. Asus, Gigabyte, etc. etc select one profile, the unlimited one, call it auto-optimize and they're done. Most of the time, everything works fine because the intel can regulate itself and it is built to handle the power and heat. Worst thing that happens is it throttles itself when it runs too hot.

The tampering of i7 and i9 CPUs had to do with how the CPU was maniputaled to avoid throttling, and how performance was eeked out of it at the expense of stability.
Read the damn articles written about this, not just the comments where you pick fights with anyone you don't agree with.

How Intel designed their overclocking utility is completely unrelated.
The auto-optimize setting they choose is unrelated.
You waving it like some smoking gun shows how little you understand.

 

[YSCCC]

They are the best CPUs, the most efficient CPUs and least power hungry CPUs. Just because the i9 can take 350w doesn't mean it has to.

As much as I am satisified with the performance of my now regarded as "extreme" profile 14900K running stable (touching wood), but it is by no means most efficient and least power hungry... at gaming and photoshop raw photos converting tasks, it eats up some 30-50W more than my friends using Ryzen 7950x, and it also produce a ton of waste heat which struggling my Noctua NH-U12A tower cooler as well. This is one of the overall fastest CPU in the market right now, but no way it can be tied with power efficient and least power hungry

 

[TerryLaze]

That's because you are arguing from the position of ignorance.

As I said, you have no idea what you are talking about.

Exhibit A:

https://imgur.com/a/ASzJyAb

View: https://imgur.com/a/ASzJyAb


I set PL1 and PL2 limits for my Xeon w5-2455X to 300 W and 360 W in BIOS (up from defaults of 200W and 240W). So why are they both set to 500 W now you ask?

In XTU there is a button front center (in simple view, not advanced view mind you!) which says "Optimize Power & Current Limits". Out of curiosity I clicked it, and you can see the result on the image.

Yes, Intel XTU "optimized" my PL1 and PL2 to 500 W!

So please stop with the BS about Intel not pushing unsafe values themselves.

Or you know, they could sell tighly binned chips at a higher price with higher overclocking limits and potential which are UNLOCKED BUT NOT UNLIMITED.

Exhibit B:

https://imgur.com/a/lrxiMfU

View: https://imgur.com/a/lrxiMfU


Can I really run this CPU at 16 GHz SAFELY?!?

If the "optimized" power is 500 W for meager 4 GHz would they set 2,000 W for 16 GHz?

It's a rhetorical question, hope it proves the point which was -- if Intel wanted safe limits they would limit what values can be set in the CPU itself to begin with. Those values could've still been considerably higher for K SKUs compared to non-K SKUs, BUT STILL WITHIN SPEC.

Then I guess I am lucky that in the case of my CPU Intel XTU has exceeded spec by only 2.5x, right? 🙄

You sound more and more like an Intel fanboy defending them. Multi-threading scenario loses 20% of performance. Most loads nowadays are multi-threaded so everything is affected.

Again you have no idea what you are talking about. When you install a CPU on a mainboard for the first time you are asked to enter BIOS to configure it. If you ignore that it's on you.

No, Intel can be blamed for allowing setting of unsafe values which are out of spec even for unlocked CPUs and for their XTU which also "optimizes" stuff to a crisp if allowed.

The first thing that happens when you click on any tab in XTU that could change settings beyond specs is for this window to pop up.
You have to agree and once you agree you have agreed that you know that what you are doing can be potentially dangerous for your CPU.
That's all this whole post is about as well, agree once and you can put in any profile you want but you have to be made clear that you are not running "normal" or "default" or "stock" or "whatever" anymore.
If you are using someone else's PC or you don't remember way back when isn't intel's fault, it's yours for not remembering/knowing.
https://www.notebookcheck.net/Intel-Extreme-Tuning-Utility-XTU-Undervolting-Guide.272120.0.html

Edit:typos

 

[TerryLaze]

As much as I am satisified with the performance of my now regarded as "extreme" profile 14900K running stable (touching wood), but it is by no means most efficient and least power hungry... at gaming and photoshop raw photos converting tasks, it eats up some 30-50W more than my friends using Ryzen 7950x, and it also produce a ton of waste heat which struggling my Noctua NH-U12A tower cooler as well. This is one of the overall fastest CPU in the market right now, but no way it can be tied with power efficient and least power hungry

But you and your 14900k have the choice while your buddy is forced.
You can put in a power limit of 125W and be a fraction slower than the 7950x while using less power than he uses or you can be a fraction faster and use a lot more power than him.
Choices...now apparently a bad thing...
https://www.techpowerup.com/review/...ke-tested-at-power-limits-down-to-35-w/2.html

 

[35below0]

As much as I am satisified with the performance of my now regarded as "extreme" profile 14900K running stable (touching wood), but it is by no means most efficient and least power hungry... at gaming and photoshop raw photos converting tasks, it eats up some 30-50W more than my friends using Ryzen 7950x, and it also produce a ton of waste heat which struggling my Noctua NH-U12A tower cooler as well. This is one of the overall fastest CPU in the market right now, but no way it can be tied with power efficient and least power hungry

That's fair.
Depending on how it's tested or the intended workload it may not be the optimal CPU.

Even though i made a blanket statement it was in different context. Regarding intel motherboard tests/reviews, and the Raptor/Alder lineup, the i9s are the "best".

But for example, i would not and did not consider an i9 at all. Even though it may have been the ideal choice. I decided an i5 will more than do. Now, an i9 might do everything the i5 does, and do it more efficiently, and have a longer useful lifespan. But that is only when comparing the CPUs in isolation. An i9 does need a better motherboard and a better cooler... Technically since i use the Noctua NH-D15, maybe it would have been enough.

As for best CPU overall, that claim has been made and the benchmarks and statistics back it up. But in real life scenarios, i'm not sure it fits.

Since you use it, you probably have a better idea. As to your cooler, i don't think Noctua recommends it. It seems the NH-U12A should struggle with an unlimited 14900K.

Choices...now apparently a bad thing...

In trhe context of this topic, the choice to overdrive CPUs was made silently.

As for tweaking power settings manually, maybe that's a good thing. In my opinion most users would rather set it all to auto and trust the motherboard, CPU and firmware know what they're doing.

 

[TerryLaze]

In trhe context of this topic, the choice to overdrive CPUs was made silently.

As for tweaking power settings manually, maybe that's a good thing. In my opinion most users would rather set it all to auto and trust the motherboard, CPU and firmware know what they're doing.

That's what I'm saying, intel forcing mobo makers to use a baseline setting as default will enable the end user to have the choice of what they will do instead of being forced to run a CPU at unstably high overclocks.

Because in my mind this will translate in all future bioses having a splash screen at first boot up telling the end user that the preselected profile is safe but also giving them a list of other profiles if they want to try for higher performance, either right away or at any time by going into the bios.

 

[CmdrShepard]

The first thing that happens when you click on any tab in XTU that could change settings beyond specs is for this window to pop up. You have to agree and once you agree you have agreed that you know that what you are doing can be potentially dangerous for your CPU.

That doesn't change the fact that Intel's OPTIMIZED settings are 2.5x out of CPU spec.

That's all this whole post is about as well, agree once and you can put in any profile you want but you have to be made clear that you are not running "normal" or "default" or "stock" or "whatever" anymore.

So you mean like this message shown by every BIOS since forever?
https://softwareg.com.au/blogs/computer-hardware/new-cpu-installed-press-f1

 

[TerryLaze]

That doesn't change the fact that Intel's OPTIMIZED settings are 2.5x out of CPU spec.

It does change the fact that you know about the risks though, at least you clicked on the agreement stating as much.

I mean you claim that this option is front and center and not in advanced but your proof of it is showing the advanced tab.

If you are talking about this....being able to read helps a lot with trying to use a PC... OVERCLOCKS is the sixth word in the description, it's not THAT hard to figure out that it might be overclocking your CPU.

So you mean like this message shown by every BIOS since forever?
https://softwareg.com.au/blogs/computer-hardware/new-cpu-installed-press-f1

Yes like this only completely different, since it would apply a default setting and allow you to also choose a different setting, without having to know anything about bios settings.

 

[CmdrShepard]

If you are talking about this....being able to read helps a lot with trying to use a PC... OVERCLOCKS is the sixth word in the description, it's not THAT hard to figure out that it might be overclocking your CPU.

Except that I am not talking about Speed Optimizer at all...

Exhibit C:

https://imgur.com/a/aGmBnLE

View: https://imgur.com/a/aGmBnLE


That's the default view in which Intel XTU launches when first installed (after you see the disclaimer).

There's a big button "Optimize Power & Current Limits" there front and center.

It totally doesn't optimize them -- it pushes them to 2.5x of the default PL1 and PL2 on my CPU, and pushes the Icc MAX to 500 A.

Exhibit D:

https://imgur.com/a/o5UnKQ5

View: https://imgur.com/a/o5UnKQ5


Nowhere does it mention OVERCLOCKING in the explanation of what it does, and from the text it isn't clear whether it is going to reduce or increase the values mentioned.

Finally, "restore system defaults" restores whatever values were set in BIOS by me or mainboard vendor, not fused defaults for the CPU installed so there -- they couldn't even get that part right. But it's all mainboard vendors' fault because you are Intel fanboy and big benevolent Intel is blameless and can do no wrong.

 

[YSCCC]

That's fair.
Depending on how it's tested or the intended workload it may not be the optimal CPU.

Even though i made a blanket statement it was in different context. Regarding intel motherboard tests/reviews, and the Raptor/Alder lineup, the i9s are the "best".

But for example, i would not and did not consider an i9 at all. Even though it may have been the ideal choice. I decided an i5 will more than do. Now, an i9 might do everything the i5 does, and do it more efficiently, and have a longer useful lifespan. But that is only when comparing the CPUs in isolation. An i9 does need a better motherboard and a better cooler... Technically since i use the Noctua NH-D15, maybe it would have been enough.

As for best CPU overall, that claim has been made and the benchmarks and statistics back it up. But in real life scenarios, i'm not sure it fits.

Since you use it, you probably have a better idea. As to your cooler, i don't think Noctua recommends it. It seems the NH-U12A should struggle with an unlimited 14900K.


In trhe context of this topic, the choice to overdrive CPUs was made silently.

As for tweaking power settings manually, maybe that's a good thing. In my opinion most users would rather set it all to auto and trust the motherboard, CPU and firmware know what they're doing.

It isn't the optimal CPU for most cases, not by any meaningful margin from the AMD solution. I did change to this is because I used to upgrade from an Sandy bridge i7 2600k directly to a 12700KF and a gigabyte Z690 MB, now after more than a year, the CPU have not great IMC which don't allow me to run my pair of gskill 64GB ram with XMP, and that playing certain addon with flight sim requires a better CPU with an IGPU also, so, the most budget friendly solution to me was to just update the bios and get an i9, the NH-U12A is actually doing fine with the intel extreme profile which I found strikes best power/performance ratio for me.

But that being said I should have gone AMD back then and not hoping intel will be more stable and runs games+photoshop faster from old habit. the whole 12th and 14th gen issues is really the craziest since I've started building PC from the Athlon and Pentium 3 generation. 12700kf took over half a year since release tot run DDR5 stable without crashing left and right, intel goes from an i7 being able to run fine on boxed cooler to basically requires a NH-U12A or better, I won't have hope the now i9 could last anywhere near the old sandy bridge does, even games well with a 3070 for a lot of cases