On Tue, Feb 06, 2024, Thomas C. Bishop via AMBER wrote:
>I"m having some difficulty figuring out amber's GPU performance on 3
>different machines in my lab. Something's not right.
>NOTE the RTX 3060 TI is a LHR version.
>
>By the specs the RTX3060TI should be about 1x to 2x faster than the
>RTX2060 Super but it is gets about 1/2x the ns/day.
>And the RTX4090 does not seem to get expected speeds regardless if I
>compare to the 3060Ti or 2060 Super.
>`
>The system has 47407 atoms and all simulations use
>gbsa=3 :Compute surface area using a fast pairwise approximation
> suitable for GPU machine.
I don't think any of the reported benchmarks use gbsa=3, so it may be hard to
know what the "expectations" should be.
I'd recommend running at least one of the standard GB benchmarks on your
hardware, to see if the relative timings you get match what the benchmark
authors expect. (Or, try your system without gbsa=3). That will help
narrow down possible origins of the behavior you describe.
....dac
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun Feb 11 2024 - 08:30:02 PST
