News Overclocked Core i9-12900K Hits 5.2 GHz At 330W, Challenges Ryzen 9 5950X

Makaveli

Distinguished
Jan 15, 2001
1,170
236
19,670
4
330w, vs the 5950x at 105w
As pointed out you are comparing overclocked to stock.

a 5950X uses more than 105 Watts when you also overclock it I think it hits about 230watts which is still 100 watts lower than this and that is only 8 cores for intel and 16 for AMD so not really a good look. If intel was able to do 16 Performance cores in this thing it would do like 450watts lol.
 

spongiemaster

Respectable
Dec 12, 2019
1,488
741
2,060
0
a 5950X uses more than 105 Watts when you also overclock it I think it hits about 230watts which is still 100 watts lower than this and that is only 8 cores for intel and 16 for AMD so not really a good look. If intel was able to do 16 Performance cores in this thing it would do like 450watts lol.
What is the power consumption for a 5900x (or 5800x) overclocked to the single threaded performance level of the 12900k here and the multithreaded performance level of a 5950x?
 
Reactions: artk2219

Phaaze88

Titan
Ambassador
Let me just see the power consumption while using the blasted thing as Intel intended, sheesh.
The cpu is likely pretty darn fast and power efficient without that nonsense.
Overclocking is dead The bar for overclocking raised some time ago, cutting off the more casual users.
 
Reactions: alceryes

Makaveli

Distinguished
Jan 15, 2001
1,170
236
19,670
4
What is the power consumption for a 5900x (or 5800x) overclocked to the single threaded performance level of the 12900k here and the multithreaded performance level of a 5950x?
That is impossible to know unless you have LN2 and even then still a maybe.

Zen 3 doesn't clock high enough with standard cooling to answer that for ST.

A 5950X with PBO already passed this score in Multi and that is using 200-230 watts.
 

spongiemaster

Respectable
Dec 12, 2019
1,488
741
2,060
0
That is impossible to know unless you have LN2 and even then still a maybe.

Zen 3 doesn't clock high enough with standard cooling to answer that for ST.
Exactly. People dumping on these results are missing the forest for the trees. What people should be looking at is the performance, not the power usage. It doesn't matter how much power you dump into a Zen3 CPU, you aren't going to match the single threaded performance of the linked 12900k results. Then you have to look at the flip side. If Zen3 can't reach this level of performance, then to level the playing field, you have to slow down the Intel CPU. If you lower the clock speeds of the 12900k to reduce performance 24% and equal Zen3 results, then how much power is the 12900k using? I bet we'll find the efficiency is pretty similar at that point.

A 5950X with PBO already passed this score in Multi and that is using 200-230 watts.
Again, missing the bigger point. A 16 core/32 thread AMD CPU is losing to an Intel CPU using only 8 hyper threaded large cores and 8 smaller non-hyper threaded cores. 24 threads are beating 32 threads. The leaked specs of the 12900k have shown a 228W PL2 and all core boost of 5Ghz. That would mean that this particular 12900k overclock took a massive 102W (45% higher)to increase clock speeds of the 8 large cores by only 200Mhz (4%). That's a horrific trade off that no one should be doing on an every day PC. Knock less than 4% off the 12900k multi threaded results and then compare 228W to 200-230W, and AMD still wins, but it is no longer a blood bath in the efficiency comparison. No matter how much power you use, you're not going to be able to overclock a 5900x with 12 large cores to equal the performance of a 5950x. Up to 16 threads and Alder Lake is going to have a pretty significant lead vs Zen 3 which is what should matter to most people here but is getting missed by the crowd ROFL'ing at the 330W headline.
 
Last edited:
Again, missing the bigger point. A 16 core/32 thread AMD CPU is losing to an Intel CPU using only 8 hyper threaded large cores and 8 smaller non-hyper threaded cores. 24 threads are beating 32 threads.
Small cores where disabled in this O/C bench.
Unless this means that you can OC only the p cores whiles the e cores keep running at stock.
But this is only one bench and cpu-z at that (super short bench) so it's pretty useless for gauging normal performance.
A Chinese enthusiast this week overclocked his Core i9-12900K to 5.20 GHz on all high-performance cores and managed to leave behind AMD's Ryzen flagship in a multi-threaded test.
 
is single core performance still ok to something we look today?
In cinebench or cpu-z? no, completely useless since you are never going to use this type of software on only one core, if you would even use them at all.
Sadly serial execution which would be useful in showing you how well every day normal stuff would run has been completely removed from every benchmarking suit to make people by more cores.
 
Reactions: artk2219
Oct 5, 2021
18
3
15
0
Exactly. People dumping on these results are missing the forest for the trees. What people should be looking at is the performance, not the power usage. It doesn't matter how much power you dump into a Zen3 CPU, you aren't going to match the single threaded performance of the linked 12900k results. Then you have to look at the flip side. If Zen3 can't reach this level of performance, then to level the playing field, you have to slow down the Intel CPU. If you lower the clock speeds of the 12900k to reduce performance 24% and equal Zen3 results, then how much power is the 12900k using? I bet we'll find the efficiency is pretty similar at that point.


Again, missing the bigger point. A 16 core/32 thread AMD CPU is losing to an Intel CPU using only 8 hyper threaded large cores and 8 smaller non-hyper threaded cores. 24 threads are beating 32 threads. The leaked specs of the 12900k have shown a 228W PL2 and all core boost of 5Ghz. That would mean that this particular 12900k overclock took a massive 102W (45% higher)to increase clock speeds of the 8 large cores by only 200Mhz (4%). That's a horrific trade off that no one should be doing on an every day PC. Knock less than 4% off the 12900k multi threaded results and then compare 228W to 200-230W, and AMD still wins, but it is no longer a blood bath in the efficiency comparison. No matter how much power you use, you're not going to be able to overclock a 5900x with 12 large cores to equal the performance of a 5950x. Up to 16 threads and Alder Lake is going to have a pretty significant lead vs Zen 3 which is what should matter to most people here but is getting missed by the crowd ROFL'ing at the 330W headline.
Finally..someone who gets it! Bravo!
 
Reactions: artk2219

VforV

Respectable
Oct 9, 2019
375
189
1,870
1
Nice FAIL intel.
The more I see about this Alder Lake CPU, the less I am impressed.

Horrible efficiency, barely faster, 1 year later, very expensive (CPU+MB+DDR5) and needs help from Win11 and DDR5. Those are a lot of minuses... I don't like it one bit.

Also, does no one see this?
11900k scores higher than 5950X in CPU-Z and other synthetic benchmarks, but gets beaten in gaming benchmarks.

So the actual lead of Alder Lake in gaming benchmarks will be much smaller than it looks in these synthetics. Not to mention if you take away Win11 and DDR5, I expect it will not win then anymore.

And then comes Zen 3D. Yeah, Alder Lake is a FAIL for me, if these leaks are true.

Even if I was an intel fanboi I would skip this 1st gen and wait for Raptor Lake which will have matured tech and process, better DDR5 and PCIe 5 prices and options and support, matured Win11 drivers, etc.
Everything looks better for Raptor Lake, but not only because it's the next step, but because of external factors too. This Alder Lake is like Zen1, but even less impressive somehow. Meh...

The only thing I like about Alder Lake is that it pushes AMD to do better and lower prices (hopefully).
 
Reactions: artk2219
Also, does no one see this?
11900k scores higher than 5950X in CPU-Z and other synthetic benchmarks, but gets beaten in gaming benchmarks.
The only game bench we have so far for alder lake shows an almost 40% performance advantage over the 5950x.
https://www.tomshardware.com/news/core-i9-12900k-destroys-ryzen-9-5950x-38-percent-ashes-of-the-singularity
The Core i9-12900K outperformed the Ryzen 9 5950X by 38.6%. Both systems were using version 3.10.191346.0 of the benchmark tool so it should be a fair apples-to-apples comparison.
 
Reactions: artk2219

spongiemaster

Respectable
Dec 12, 2019
1,488
741
2,060
0
Small cores where disabled in this O/C bench.
Unless this means that you can OC only the p cores whiles the e cores keep running at stock.
But this is only one bench and cpu-z at that (super short bench) so it's pretty useless for gauging normal performance.
From what I have read the small cores are stuck at 3.7Ghz, and don't overclock when you overclock the P cores. There is no way 8 Golden Cove cores overclocked by 200Mhz beat a 5950x in a multithreaded benchmark unless the benchmark is completely bogus. Not even Intel is trying to claim that level of performance uplift from Golden Cove. 18% IPC gain over Tiger Lake does not equal twice as fast as a Zen 3 core.
 
Reactions: artk2219

spongiemaster

Respectable
Dec 12, 2019
1,488
741
2,060
0
Holy cow that is a lot of power if true...

At that power level may as well just use ThreadRipper for the extra wideness and more memory channels which should still edge out DDR5 at launch. Maybe that's why Intel can't price this thing higher? Hm... Also, TR 5K around the corner, supposedly with the 3D-cache is going to make Alder Lake sweat profusely during it's prime time. That's not even counting Ryzen with the 3D-cache refresh, which I think it's going to be low volume. More like a "we're still here" from AMD, haha.

Day 1 reviews are going to be saucy.

Regards.
 
Reactions: artk2219
From what I have read the small cores are stuck at 3.7Ghz, and don't overclock when you overclock the P cores. There is no way 8 Golden Cove cores overclocked by 200Mhz beat a 5950x in a multithreaded benchmark unless the benchmark is completely bogus. Not even Intel is trying to claim that level of performance uplift from Golden Cove. 18% IPC gain over Tiger Lake does not equal twice as fast as a Zen 3 core.
Yeah that makes much more sense, which is why I included it as an option, it just wasn't clear from the article.
It's actually quite relieving that you will be able to overclock without having to shut off the small cores.
40% advantage over a 5900x, not 5950x. The benchmark was updated to support 24 threads, so 1/4th of the 5950x was idle during the benchmark, making it a 5900x.
Sure but that doesn't make the 5950x any faster in that bench,
and if you compare the 11900k to the 5950x in game benches as V for V did a CPU that has twice the cores and is $150 more expensive then you can do this 12900k vs 5950x comparison in ashes as well.
 
Reactions: artk2219

ASK THE COMMUNITY