I didn’t link to them, not for years now.

If you don’t believe in phoronix methodology, why link to them in a blog post?

You are not going to get diligent investigative testing from phoronix, nor should you draw substantial conclusions from their benchmarks.

So I wonder what linux did in the last two years? But also dfly uses a copy-on-write file system which gives history etc. Out of the box linux doesn’t have that. But overwrite is obviously always going to be much faster (and less useful) than copy-on-write

The phoronix tests don’t account for these things. Does ubuntu turn off debugging in their kernel for instance?? (Though -g -O3 can appear together)

For myself, realtime audio seems to work much better (fewer glitches/lower latency) on dfly than a low-latency version of linux (and I spent months optimizing linux!!)

But is that due to OSS/ALSA or SMP or what? Well, to answer that question, I will rely on diligent investigative testing from phoronix

https://www.phoronix.com/scan.php?page=article&item=dragonfly-55-threadripper&num=1