fieldframe wrote: ↑06 Dec 2020
So the M1 + Rosetta is clearly capable of rivaling a 2019 Intel 8-core in DSP tasks, and should be able to do considerably more than a 2018 6-core, but something is holding it back. Has anyone tried with and without hyperthreading enabled?
Youtubers already saiyng that M1 is revolution. But in music productio - better not look at not deeper tests by youtubers, but make own deeper tests to understand how really M1 work today. Running the same projects on various machines.
And looking not only to the quantity of one-plugin instances... Youtubers make tests of instances quantity at all.. and I think its bad tests to understand what M1 can do really..
And ofcourse - you can listen before - about DAWBench?
Tests used - DAWBench, its very famous real test of DAWs. And it been used in tests. (Youtubers not going this way and just run instances to max, its absolutely synthetic tests, out of the real work).
Does anbybody from youtubers try to use DAW bench? No)))) But they talk M1 is super powerful.. No.. Real tets show anothe for native M1 apps (without rosetta!)
I already write there, that we can test anything that work without Rossetta2 on arm-mac.
My example is a REAPER M1 app. With internal M1-optimised plugins. Its all. Optimised for M1.
What we can do with Reaper?
Two mac mini: 2018 and M1.
- Intel.png (200.33 KiB) Viewed 2468 times
- M1.png (221.45 KiB) Viewed 2468 times
Checking in Reaper for this two namchines
Longest RT block : 2.77/2.67 ms (hold)
Final results:
INTEL: 344 PLUGIINS
M1 : 200 PLUGINS
Finaly, Intel-Mac Mini 2018 x1.72 more powerful than M1.
Ok, Intel Mac 6-core, M1 - 4core. Anyway, per-core, its nearest results. But at the end, intel-Mac in native M1 app tests, going after Intel-Mac Mini 2018
No needs to wait Reason M1 version, to understand that M1 not better in DSP levels than Intel Mac Mini 2018 6-core.