CedricM
Did you bench Sonar Platinum and did it give the same results as Reaper ?
It seems Reaper is very strong at not clicking and popping until 100 % cpu useage, whereas Ableton encounters glitches with much more modest cpu loads. See https://www.youtube.com/watch?v=eIoAlEV2IK0
Would be interesting to know where Sonar stands.
My point in benchmarking wasn't to rank Sonar's multi-core performance...
Rather... to compare the hardware itself (mostly focusing on how Ryzen 1800 compared)
So I kept the stress-test consistent across hardware platforms.
Reaper is certainly well optimized for multi-core CPUs.
If you're talking about socket 2011-3 CPUs, yes... you can push them up to 100% load (playing audio glitch-free) at a 48-sample ASIO buffer size.
With socket 1151, you can't push the CPU load quite as far (~95%).
With Ryzen 1800x (RAM running at 3200MHz), you can push CPU load to ~93%.
Sonar multi-core performance isn't quite as efficient as Reaper... but certainly more so than Ableton Live.
One thing to keep in mind; Live was designed to work in a much different way than typical "linear" style DAW applications. ie: Audio warping is enabled by default. This is going to affect CPU use.
Even with a 6850k, you top out at ~80% load (at a 48-sample ASIO buffer size).
I don't see Live as a replacement for more traditional type composing/recording applications... so (at least for me) it's not much of a real-world limitation.
The complexity of the audio signal has no affect on CPU load.
Doesn't matter if you're feeding the plugin a sine wave, white/pink noise, vocals, guitar, drums, etc.
CPU load is consistent.
I agree that DAW Bench isn't 100% analogous to a typical real-world project... in that you're not looking to run 500 instances of the same exact plugin.
It's just a relatively simple way to mark/compare audio specific low-latency performance.
An easy point of comparison...