• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

Another Day, Another Intel Core i9-12900K Benchmark Leak

Until I see cinibench scores I'm not making any judgements but these leaks of cpu-z are in no way a reflection of real load on all cores.

The Cinebench score leak was last week.

In case you missed it:

1632919912770.png

And since the finger-marker blocking the specs can be easily reversed by changing the contrast on the image it's running on a Gigabyte Z690 AORUS ULTRA.
 
Great, but i bet the power usage is as good as twice compared to a Ryzen.
 
The Cinebench score leak was last week.

In case you missed it:

View attachment 218693

And since the finger-marker blocking the specs can be easily reversed by changing the contrast on the image it's running on a Gigabyte Z690 AORUS ULTRA.
The finger marking isn't really covering anything interesting I'm more confused with the results and why the table only shows first gen ryzen's and years old threadrippers and task managers Cpu utilisation only showing it using one core on a multicore task.
I'll be waiting till some real benchmarks come out this seems way too Dodgy to believe.
 
The finger marking isn't really covering anything interesting I'm more confused with the results and why the table only shows first gen ryzen's and years old threadrippers and task managers Cpu utilisation only showing it using one core on a multicore task.
I'll be waiting till some real benchmarks come out this seems way too Dodgy to believe.

Most are the default comparison results that R23 comes with. The task manager window is pretty clearly just there to show the active threads, not the utilization of the run. It was opened after the run completed almost assuredly.
 
Single core scores in CineBench R20 benchmark (from Hardware Unboxed review of the 11900K) and scaling the 12900K from the CPU-Z results;
That's just not how it works. Different achitectures scale differently across different workloads. Even in Intel's own marketing slides there were workloads Alder Lake had worse IPC in than Rocket Lake.


Besides we've already seen leaks of 800+ scores in CB.
 
That makes more sense sorry preoccupied with other things even so till consumer grade chips hit the market and we get some undoubtedly unbias benchmarks I'm not making any judgements.
 
Well I think CPU upgrades, especially in gaming & consumer rigs are not in need of revisions any faster than 5-7 years time. Especially if you're already on the top of the stack. You just barely notice it...
I agree with you. This has been my cycle now for the last 3 CPU upgrades (i7 920 > i7 5820K > 5950X). More noticeable than the increased single thread cpu performance are the native features you get with motherboards (new USB standards, M.2, ect). Not since the days of single core Athlon's have I considered upgrades within the same socket type.
 
Alder Lake is shaping up to mostly be what I expected it to be. In some tasks the single core performance is going to be greater but the multi core crown will still go to Zen 3. My 5900X scores just above 700 in single core but scores 10400 in multi core. While the 12900K (if these numbers are correct and if they are on a system not overclocked on an industrial chiller) outscores my 5900X, the 12900K is going up against the 5950X (as they are both 16 core components). The 32 thread 5950X will still hold a performance advantage in multi core applications over the soon to be released Intel flagship 12900K. It will be interesting to see if the impressive single core scores (if legit) will translate into gaming performance or if the new arch will be hamstrung with a hefty latency penalty. It will also be interesting to see what AMD can bring to the table with a Zen 3 refresh. It is easy to forget that the 5900X and 5950X are last years flagships.
 
It's known that this performance comes at the cost of huge power consumption and thermal output, according to multi sources, and this is recently confirmed by what's just posted by Toppc.
I wonder what's got into Intel's mind. Why can't they bring the power consumption down, even though they have used 10 nm process node... no... Intel 7 process node...
I think they could totally bring power consumption down by dialing back the boost clocks. They likely designed this chip for a smaller node than it's getting. Even with Rocket Lake, the architecture largely behaved until you tell the motherboard to ignore the power limits. I think these current designs aren't good enough to compete until power consumption is ignored. I'm going to go out on a limb and guess that the 12900K is going to be supply limited, just not for review sites.
 
Lol intel has 90% of the market they not worried
If you follow their press I'm not sure how you can possibly draw that conclusion the past few years.

They've been losing share and trying their hardest to mitigate this, which only works in part simply because they can keep supply chains filled. But even that wasn't always the case, I'm sure you remember.

If you're not worried, the last thing you do is keep trying to save face on your 10nm debacle and the absolute last thing you do is rearrange your chips to make them work on an unplanned node (14nm). Their stacks have been a complete mess, roadmaps got thrown upside down faster than we could read the articles.

The fact is, we don't even know what Alder Lake will be doing in real world and if its not going to deliver, which is a pretty high bar to meet right now, the above situation has not ended, it has worsened for Intel.
 
Back
Top