Replies: 2 comments
-
You should try -r16 -K 2 and report back |
Beta Was this translation helpful? Give feedback.
0 replies
-
-K multiplies the -r value for phase 2. Phase 2 seems to have the fastest time when -r = logical cores, but phase 3 (dominant for time) is fastest with -r = physical cores. -K 2 allows both phase 2 and 3 to operate at their fastest -r settings simultaneously. It's not a huge gain, but it shaved an extra minute off of my previously fastest settings. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
I've seen in the update we now have:
-K, --rmulti2 arg Thread multiplier for P2 (default = 1)
What does this do exactly? I'm wondering whether it would benefit me to use it but I'm not sure how. I have 2x E5-2640v3 processors giving total 16C/32T and I'm running -r 32 without trouble, making plots in 45min
Beta Was this translation helpful? Give feedback.
All reactions