16-inch MBP 2x slower than M1 MacBook Air in a real-world Rust compile

Re:16-inch MBP 2x slower than M1 MacBook Air in a real-world... by w0utert on Wednesday November 18, 2020 @05:33AM (#25135214)

Yes but the point is that Apple doesn't need to scale the M1 up to desktop frequencies, because it already is faster than x86 in single-threaded workloads, at lower clocks and significantly lower power. To scale up the multithreaded performance they just have to increase the core count and scale the cooling system accordingly, ie: exactly like you would have to for x86. A decent desktop cooler can dissipate enough heat to run CPU's with 100+ Watt TDP's, while the M1 in the current Mac Mini sits around ~20W estimated if you discount the RAM.

So again, what would make anyone think that an M1 with decent cooling would not be able to maintain the current ST performance indefintely, or a hypothetical 8+8 or even 16+16 core M1X or M2 with a TDP of 100W and top-notch cooling solution would be impossible?

[ Parent | on HN ]

Re:16-inch MBP 2x slower than M1 MacBook Air in a real-world... by spronkey on Thursday November 19, 2020 @05:52AM (#25148202)

Multicore scaling on M1 doesn't appear to be as efficient as that on Ryzen - for example, the 4200U is able to beat the M1 in multicore tasks at quite a similar power draw, but gets soundly beaten in single core. Ryzen's big advantage over previous multicore implementations was infinity fabric - so clearly there's more to scaling than just the actual compute cores.

Don't forget that scaling up is also not just about frequencies, there are also packaging considerations - the CPU dies have to actually be able to dissipate the heat generated, and the package itself has to be able to do so as well. I'd expect that this is something AMD and especially Intel have a leg up over Apple with - although, considering they've already tread that ground it makes Apple's job a bit easier too.

[ Parent | on HN ]