# Intel Core i9-12900K

#### TheoneandonlyMrK

I'm more concerned with the heat output in one room. Though I'm in central texas so it's a bigger concern for me than others. (my solution was a mini-split in my server room and switching to a minipc at my desk where I remote to my gaming system. Stays cools and I don't care how much heat it generates.)
I live in England and I would worry about the heat!,

12900k is totes pointless IMHO, the 12700K is pretty good the 12600K is good in its bracket but all in not worth most people upgrading to If your on the last generation and casual gaming.

#### Valantar

all in not worth most people upgrading to If your on the last generation and casual gaming.
There is essentially no scenario where upgrading makes sense if you're on the last generation, period. The gains are never that significant.

#### RandallFlagg

Supporter
I have my doubts. Because this is the first x86 product like this on Windows, and the hybrid approach is limited to only part of the 12 series, it means 99% of the hardware out there will still be homogeneous CPU architecture, and for many years to come with the way our hardware can now last for so long. It’s going to be on Intel to make this work, then MS, and maybe developers will jump in. I could easily see developers just saying “use different hardware” if you have issues, at least for a while.

Anandtech came to the conclusion that DDR5 does contribute to the performance increase. They have 2 pages of DDR4 vs DDR5 that show measurable gains. It’s not across the board, but significant, especially mutlithread. They even concede that AMD should see similar gains when they implement DDR5, though they are obviously further out.

That would be because AnandTech uses JEDEC standard RAM. This is garbage to any DIY builder. The benchmarks are also being run on Windows 10 for inexplicable reasons.

I will say I agree in the future DDR5 will be faster, but it doesn't compete well right now against enthusiast grade DD4 on Alder Lake (i.e. anything over 3600).

To trade blows with moderately fast DDR4-3800 (basically $200 for 32GB), you would need top shelf DDR5-6000+ (which is like$800 and effectively unobtainable).

Case in point :

#### Lycanwolfen

Actually Windows 10 support Big and Small cores always has but Microsoft turns it off. The Key is HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Power\PowerSettings\54533251-82be-4824-96c1-47b60b740d00\93b8b6dc-0698-4d1c-9ee4-0644e900c85d Change the Attributes to value of 2 and also HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Power\PowerSettings\54533251-82be-4824-96c1-47b60b740d00\bae08b81-2d5e-4688-ad6a-13243356654b Change the Attributes to value of 2. Also a golden oldie which allow much better boosting in WIndows 10 HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Power\PowerSettings\54533251-82be-4824-96c1-47b60b740d00\be337238-0d82-4146-a960-4f3749d470c7 and change Attributes to 2.

Once those are at 2 goto power settings and advanced under processor you can now access those features.

#### HenrySomeone

There is essentially no scenario where upgrading makes sense if you're on the last generation, period. The gains are never that significant.
Well, I guess if you were a really, really hardcore AMD fanboy with an FX 8370 who wouldn't touch Intel with a 10 foot pole, then I guess moving to Zen1 kinda, sorta made sense?

#### W1zzard

Staff member
- Did you use a U14s or a U12s for overclocking? @W1zzard
meh.. fail .. NH-U14S of course, review has been updated. Thanks!

Very good work but if you allow me I can find out the reason why you changed the Zen setup from EVGA X570 DARK with 4000mhz@2000 IF memories that you used in the last reviews you did on MSI X570 and 3600@1800 IF memories?
The EVGA Dark is for in the graphics card test system. Not the CPU review system. None of my "CPU review" Zen 3s can do 2000 MHz

#### Valantar

Well, I guess if you were a really, really hardcore AMD fanboy with an FX 8370 who wouldn't touch Intel with a 10 foot pole, then I guess moving to Zen1 kinda, sorta made sense?
That's not a generation, that's a whole new species

#### docnorth

Anandtech came to the conclusion that DDR5 does contribute to the performance increase. They have 2 pages of DDR4 vs DDR5 that show measurable gains. It’s not across the board, but significant, especially mutlithread. They even concede that AMD should see similar gains when they implement DDR5, though they are obviously further out.
Anandtech's review uses W10 and and today's JEDEC speeds and indeed came to the conclusion that some of the so-called multithreaded CPU tests benefit from DDR5. OTOH @W1zzard found that the hybrid architecture does not (yet) work properly with at least 4 CPU tests and 3 of them saw a huge performance increase just by disabling the E cores. So DDR5 seems to be an advantage (which will increase in a few months) for AL and the architecture's immaturity a disadvantage (which should shrink in a few months) when it comes to productivity work, but right now we don't know what has bigger impact. Anyway for MT depended work 5950x probably remains the best choice, if someone doesn't want a server.

#### AnarchoPrimitiv

So, if AMD's V-cache does achieve 15% average gaming performance uplift like they've claimed, Alder lake is handily beaten? Is that the case? How does a company with 6.5x the R&D budget and 8x the annual revenue not destroy AMD?

#### ncrs

So, if AMD's V-cache does achieve 15% average gaming performance uplift like they've claimed, Alder lake is handily beaten? Is that the case? How does a company with 6.5x the R&D budget and 8x the annual revenue not destroy AMD?
Sometimes growing too big as a company is a detriment, and efficiency is lost.

#### docnorth

Only half the story, still need to see how AMD will perform under DDR5.
Unfortunately we don't have an answer to this, so to test Alder Lake with both DDR4 and DDR5 is the next best thing to do when it comes to objectivity.

#### HenrySomeone

So, if AMD's V-cache does achieve 15% average gaming performance uplift like they've claimed, Alder lake is handily beaten? Is that the case? How does a company with 6.5x the R&D budget and 8x the annual revenue not destroy AMD?
Because AMD is using a fab process from a company with an even higher R&D budget...however the price for that is not really getting the volume to make a dent in Intel's sales and profits and Apple occupying more and more of its best nodes will only make things harder...

Because AMD is using a fab process from a company with an even higher R&D budget...however the price for that is not really getting the volume to make a dent in Intel's sales and profits and Apple occupying more and more of its best nodes will only make things harder...

#### lexluthermiester

No, it's not, DDR4 vs DDR5.

And it's not about it being unfair, having just one platform on DDR5 isn't enough to infer how good these CPUs actually are. Any CPU with faster memory will also perform better, nothing new here.
Sorry but you're missing context and it's very simple. The comparison here is the top consumer CPU from AMD and the new top consumer CPU from Intel on the given platform in which they run, along with many other previous models to compare against. The RAM type is a part of the testing and is statistically irrelevant in the scope of the testing done. It may not be perfectly apples to apples but it's as close as W1zzard(or anyone else) can get currently due to the specifications of the technology.

Just because there is an aspect of the testing done that does not meet with your satisfaction does NOT invalidate the testing as a whole. Your limited perspective is not the problem of reviewers/testers.

Last edited:

#### BorisDG

Looks like my x299 will live on another generation....
Same. It's still beast platform + CPUs. Imma replace it just with an another HEDT. I want something faster than 15% and with good thermal/power performance.

Last edited:

#### Zubasa

Sorry but you're missing context and it's very simple. The comparison here is the top consumer CPU from AMD and the new top consumer CPU from Intel on the given platform in which they run, along with many other previous models to compare against. The RAM type is a part of the testing and is statistically irrelevant in the scope of the testing done. It may not be perfectly apples to apples but it's as close as W1zzard(or anyone else) can get currently due to the specifications of the technology.

Just because there is an aspect of the testing done that does meet with your satisfaction does NOT invalidate the testing as a whole. Your limited perspective is not the problem of reviewers/testers.
There is a problem with the particular setup though, most reviewers used the G.skill DDR5-6000 CL36 sticks that came with their review kits.
Those sticks are not listed anywhere and not even up for pre-order. So basically unobtainium as of the time of release.
FYI those sticks are overclocked to the max, in fact GN could not manage to run them at all at 6000 with their particular setup.

Last edited:

#### Vya Domus

Just because there is an aspect of the testing done that does meet with your satisfaction does NOT invalidate the testing as a whole.

It has nothing to do with my criteria for anything, just because these are the only circumstances under which comparisons can be made at the moment that does not mean they also automatically become "fair", that's stupid.

Other than that, compare them all you want, that's fine. I don't care.

#### lightning70

300-350W power consumption is unacceptable. Again, inefficient piece of silicon. Despite being 10nm, the disappointment is in power consumption.

#### lexluthermiester

that does not mean they also automatically become "fair"
Other than that, compare them all you want, that's fine. I don't care.
Clearly you do or you wouldn't be making a point of it.
that's stupid.
Now where's that mirror..

There is a problem with the particular setup though, most reviewers used the G.skill DDR5-6000 CL36 sticks that came with their review kits.
Those sticks are not listed anywhere and not even up for pre-order. So basically unobtainium as of the time of release.
And? You say that like you're a stranger to product launches. We both know you're not. So, seriously with that? You say that like you're implying that such spec'd RAM will never come to market and that only reviewers are going to have it. Much like Vya, your logic is deeply flawed.

Last edited:

#### Zubasa

And? You say that like you're a stranger to product launches. We both know you're not. So, seriously with that? You say that like you're implying that such spec'd RAM will never come to market and that only reviewers are going to have it. Much like Vya, your logic is deeply flawed.
There are no ETA for Equivalent DDR5-6000 CL36 kits from other manufactures either. Also the fact that it is running at a speed that is no stable on at least some CPUs.
By your logic AMD's numbers should be run with DDR4000 1:1, because future silicon might be better?
TBH I am not blaming any of the reviewers, the fact that Intel sends out overclocked products in their review kit that is not tested stable is the issue IMO.

Last edited:

#### RandallFlagg

Supporter
There is a problem with the particular setup though, most reviewers used the G.skill DDR5-6000 CL36 sticks that came with their review kits.
Those sticks are not listed anywhere and not even up for pre-order. So basically unobtainium as of the time of release.
FYI those sticks are overclocked to the max, in fact GN could not manage to run them at all at 6000 with their particular setup.

Most sites are not using DDR5-6000. Most are using DDR5-5200 which was supplied from various motherboard vendors, and quite a few are using DDR5-4800 and some even 4400. They are also using anything from DDR4-3200 CL22 (like AnandTech) to DDR4-3800 on the older Zen / Intel platforms as well as on Alder Lake. Since these are enthusiast sites I think it is the readers responsibility to know what they are looking at. If the reader doesn't know what that type of stuff means, they should probably wait for the reviews of prebuilts from Dell Acer Asus HP Lenovo etc on CNET and PCWorld. That's not a slam, there are certainly better ways to spend time than fussing about RAM speed and all of those companies have decent machines in their lineup.

#### Zubasa

Most sites are not using DDR5-6000. Most are using DDR5-5200 which was supplied from various motherboard vendors, and quite a few are using DDR5-4800 and some even 4400. They are also using anything from DDR4-3200 CL22 (like AnandTech) to DDR4-3800 on the older Zen / Intel platforms as well as on Alder Lake. Since these are enthusiast sites I think it is the readers responsibility to know what they are looking at. If the reader doesn't know what that type of stuff means, they should probably wait for the reviews of prebuilts from Dell Acer Asus HP Lenovo etc on CNET and PCWorld. That's not a slam, there are certainly better ways to spend time than fussing about RAM speed and all of those companies have decent machines in their lineup.
There is a reason why Anandtech uses "filthy" JEDEC 3200 / 4800 ram in their reviews, those are the only spec that are guaranteed to work.
Again I am not blaming reviewers themselves on using what is supplied or suggested on their review guide.

#### lexluthermiester

There are no ETA for Equivalent DDR5-6000 CL36 kits from other manufactures either.
And? Just because there's no ETA doesn't mean it's not coming soon. Remember: Pandemic economy and chip shortage. RAM makers might be holding back release dates until they know when they will have stock. Regardless of why, the fact that reviewers were sent a certain spec of RAM does NOT invalidate the testing done. There is nothing unfair about it either. Quit your whining about a non-issue.

#### Zubasa

I am fully aware of the "Pandemic economy and chip shortage", but the fact is even the GPUs exists when they were launched.
The fact that the scalpers got hold of them means those products exists. Meanwhile not even scalpers can get their hands on these DDR5-6000 CL36 Unicorn.
Sure, brush off anything that is not inline with your views as "whining".

#### RandallFlagg

Supporter
There is a reason why Anandtech uses "filthy" JEDEC 3200 / 4800 ram in their reviews, those are the only spec that are guaranteed to work.
Again I am not blaming reviewers themselves on using what is supplied or suggested on their review guide.

I know why they use those settings, I didn't attack them for that I'm just stating a fact. I will say that their results are more of a test to see what mid level consumer grade OEM performance will be, in other words he's going to show you performance of something like a Dell Inspiron.

But who is their audience? I see in your system spec you use Team T-FORCE XCALIBUR RGB 4000 @3400 CL14. Do you think AnandTech's benchmarks using DDR4-3200 C22 are highly relevant to someone like you?