Quote:
Originally Posted by thebeatless
Or… will we get the same old spike in the middle of the cpu meter?
|
The CPU meter in PT is pretty close to worthless for this sort of measurement. Ages ago at a developer conference I had the meter explained to me. There's actually a bunch of stuff layered in there besides raw CPU load. We always end up with the lump-in-the-middle display, no matter what's really going on.
If you want an accurate look at core activity, then bring up the Mac Activity Monitor and be sure to watch the CPU history display. It displays two values for each core (because of hyper-threading) and will also show when a core is at capacity. You'll see that Pro Tools does a much better job of load distribution than the PT meter indicates. I have no idea why they persist with a meter that's so far from the truth.
NOTE: I should be clear that hyperthreading exists only in the x86 line and not in ARM (a completely different architecture, as others have said). Hyperthreading is a way the x86 gets around some complications in its prefetch and decoding queues. This is apparently not a problem in ARM, so there's no need to hyperthread. Back in my developer days, I measured hyperthreading to improve performance by no more than 30%. That's still not as good as the M1 architecture.