4 ms·

To all the commenters: This is claimed on a power adjusted basis: https://mobile.twitter.com/witeken/status/1478441037416828932/photo/1 https://mobile.twitter.c

• fouc 5y ago
  If I'm reading this right, it's saying that when M1 Max is at peak performance (at 35W), the I9-12900HK is beating this performance while at the same wattage. I'm a bit skeptical this is true
• RototRobot 5y ago
  It really wouldn't make any sense if it weren't? as by raw power the previous gen was already more powerful right?
• ksec 5y ago
  1. This is a slide from Intel. And they dont mention what sort of performance. Edit [1] 2. This is from witeken. A Well known Intel "supporter". AlderLake is good, but claiming it offer more Pref / Watt than Apple's M1 Max on a node that is generation behind would require some independent testing to verify . Because all the data known so far doesn't seems to suggest this. Especially in the case of Single Core Performance. [1] The metric used is the geometric mean of an "n-copy" SPECrate run of the C/C++ integer benchmarks in SPEC CPU 2017. You can find the M1 SPECrate 2017 results here https://www.anandtech.com/show/16252/mac-mini-apple-m1-tested/4 https://www.anandtech.com/show/16252/mac-mini-apple-m1-teste...
  • deepnotderp 5y ago
    To be fair, comparing TSMC’s nodes vs Intel’s isn’t exactly fair for CPUs since Intel has always had performance advantages beyond density such as Ion, cobalt, air gaps, etc. Edit: Ion= on state current