macrumor上已经有人指出了：光ps5那个散热片就决定了m1m没戏
光比指标是纸上谈兵
Such an illiterate logic ?.
Apple’s marketing and their fanbase never fails to impress me, especially on MacRumors…
10.4TF doesn’t mean that it can actually use 10.4TF if it’s power limited to +-60W and probably even bandwidth starved, (lack superior L1 and L2 caches, don’t use unified L3 chance, lacks a geometry engine, doesn’t support techniques such as VRS and storage APIs…)
The AMD Radeon V is actually 14.9 TF and the AMD VEGA 64 is 13.4 TF, and both are slower than the PS5 and Series X, significantly slower in fact!
Don’t fall for the fake marketing people. By no mean these MacBooks are slow or anything, but if you actually believe that it’s faster than a PS5 you have to seek help…
For consoles to consistently hit their maximum TFlops performance they actually uses these kind of heatsink.
【 在 eGust 的大作中提到: 】
: 今天看了一个视频，里面刚好跟 ps5 比了一下：
https://youtu.be/XBJ96s2n9Es?t=645: ps5 vs m1 max 带宽 448GB/s vs 400GB/s，浮点性能 10.3 tflops vs 10.4 tflops
:
--
FROM 188.192.115.*