Live data from Hacker News

Temptation of the Apple: Dolphin on macOS M1

dolphin-emu.org

1–10 of 390 posts

Re: Temptation of the Apple: Dolphin on macOS M1

#7
post #5

Incredible the perf they get out of it. Bit confused with the graph towards the end, is perf better under Rosetta than natively?!

The rosetta vs native vs 9900k vs 8559h graph?

The only game rosetta is beating native on is rogue squadron 2. Since Dolphin is a JIT, this seems to be a case of where Rosetta's JIT is smarter than Dolphin's in terms of which ARM instructions are chosen when converting from the Intel instructions than Dolphin when converting from the emulated PPC instructions.

Unless you're comparison is the 8559h and not the "native" bar. I mean, the 8559h is a mid range older Intel CPU and it's hard to understate how much Intel stagnated since Sandy Bridge (and especially since Skylake).

Re: Temptation of the Apple: Dolphin on macOS M1

#9
> There's undeniable excitement for the next generation of AArch64 hardware to see how much further that this can go.

This is what I am looking forward to. In the case for Apple Silicon, the next generation will be even better and is not far off from announcing the newer processors that will supersede M1 in WWDC.

The M1 only shows what's possible for Apple Silicon and the newer generation of ARM-based Macs will impress further. So will skip this one for now and wait what WWDC has to offer for the next generation.

Re: Temptation of the Apple: Dolphin on macOS M1

#10
post #7
post #5

Incredible the perf they get out of it. Bit confused with the graph towards the end, is perf better under Rosetta than natively?!

The rosetta vs native vs 9900k vs 8559h graph? The only game rosetta is beating native on is rogue squadron 2. Since Dolphin is a JIT, this seems to be a case of where Rosetta's JIT is smarter than Dolphin's in terms of which ARM instructions are chosen when converting from the Intel instructions than Dolphin when converting from the emulated PPC instructions. Unless you're comparison is the 8559h and not the "native…

Rosetta is faster than native in that case because the AArch64 JIT has to fall back to the interpreter for memchecks (unlike the x86-64 JIT).
Post reply on HN