What I remember from previous conversations was that some AMD chips have better single thread performance, but Intel had better multithreaded performance. It had to do with AMD interconnects between cores having higher latency than Intel’s.
Is someone up on the evolution of AMD and Intel professors, as it affects Rack multiprocessing?
Last year, I went from a 12th gen Intel i7 laptop to a Ryzen 7 7800x3D desktop PC.
A big factor for me was not to use process lasso anymore while running Ableton. If memory serves, I also turned off e-cores in VCV Rack Pro 2 with process lasso, just to be safe.
I don’t run massive patches and tend to just build playable synths and effects in VCV. I feel Ableton is much more stable now. VCV always seemed stable in either system but I’m pretty sure I never came close to really stressing it either. And FWIW, I run like 99% of the time as a plugin inside Ableton and rarely use the standalone.
According to the dawbench tests the new intel core ultra 2xx perform quiet good as well as the amd 9800x3d
And you can read/listen more here:
So you used Process Lasso to keep Rack on performance cores?
I have to use windows settings on my laptop to keep Rack on the GPU instead of built in Intel graphics, otherwise performance is slow and glitchy.
Having low power cores in the mix would be another annoyance.
I don’t know why we can’t have CPUs that just work properly without messing around.
Even the Apple M series chips don’t handle multiprocessing very well. Given the inevitable hardware limits on single-core performance, you’d think chip makers would do a better job with multi core/multi thread operation.
This is the kind of Rack issue some guidance from @vortico or other Rack devs would be welcome.