So as predicted, the E cores are primarily there to artificially inflate the synthetics (Cinebench, etc) in the MOAR CORE number-chasing p*ssing contest
artificially inflate ??
Cinebench is 3d rendering benchmark used to evaluate Cinema 4D software performance. Basically, faster cinebench mean better rendering performance
(and even then Alder Lake seems to get thrashed by the Ryzens in 7zip, Corona, Blender, etc),
12900K outperformed 5900X in blender and Corona
7zip seems that only benchmark that did not benefits from E cores (maybe bug or 7zip failed to detect E cores) but for some reason DDR5 showed massive gains in compression (even beat 5950X) which is weird
whilst gaming benchmarks are barely 9% higher on average vs the 2 generation old i9-10900K, and at least part of that 9% includes "+20% higher" IPC & faster RAM? Also, no thank you to 355w CPU's.
Gaming benchmark could be GPU limited. CPU will not magically increase fps on GPU limited scenario
There is also cheaper alternative like 12700KF and 12600KF which is not much different in gaming performance anyway.
https://wccftech.com/review/intel-c...-wifi-g-skill-trident-z5-ddr5-6000-memory/13/
355w is total system draw and not CPU alone (even Ryzen like 5900X consumed 242 watt and we know that CPU itself don't consume that much)
Also, when gaming, it seems that 12900K does not consume any more power than Ryzen 5950X or 11900K and only slightly higher than 5900X
So high power consumption happen only in highly threaded productivity benchmark and not in gaming or normal usage (if you are gamers why power consumption in blender even matter for you !!!)
Some ideas for future articles - It would be nice to see 1. The benchmarks retested with the E cores disabled just to see how much 'improvement' is down to extra cores and how much is RAM / IPC, and 2. W10 vs W11 (for those who don't want to 'upgrade' to W11 but want to see how the scheduler would impact it, eg, potential problems with heavy game threads getting assigned to the 'wrong' cores).
If E cores are useless then explain how 8 P cores destroy 11900K and 10900K (10 core) in blender, cinebench and Corona ? And even beat 12 cores Ryzen 5900X
THe IPC and clock speed of P cores is not big enough to beat 12 zen2 cores.... Only way it could beat 12 zen3 cores if the E cores is useful