- May 07, 2021
-
-
cmaffeo2 authored
-
- Apr 30, 2021
-
-
cmaffeo2 authored
-
- Apr 27, 2021
- Apr 26, 2021
-
-
cmaffeo2 authored
-
- Apr 25, 2021
-
-
cmaffeo2 authored
-
- Apr 16, 2021
-
-
cmaffeo2 authored
-
- Apr 01, 2021
- Mar 02, 2021
-
-
cmaffeo2 authored
-
-
- Feb 05, 2021
-
-
cmaffeo2 authored
Fixed groupSitee indexing, added anotehr device sync, included some (commented) printf lines for debugging
-
- Feb 04, 2021
- Jan 06, 2021
-
-
cmaffeo2 authored
-
cmaffeo2 authored
-
cmaffeo2 authored
-
cmaffeo2 authored
-
cmaffeo2 authored
Added profileStart to GrandBrownTown.cu and allowed nccl_broadcast to overlap with computeForce on GPU 0
-
cmaffeo2 authored
-
cmaffeo2 authored
-
cmaffeo2 authored
-
cmaffeo2 authored
-
cmaffeo2 authored
-
cmaffeo2 authored
-
cmaffeo2 authored
-
cmaffeo2 authored
Preliminary implementation of multi-gpu communication for nonbonded force calculation; energy kernel not yet implemented
-
cmaffeo2 authored
-
cmaffeo2 authored
-
cmaffeo2 authored
-
cmaffeo2 authored
-
cmaffeo2 authored
-
cmaffeo2 authored
Don't use warp aggregated intrinsics; the compiler for cuda9 and on can do a better job\n\nMaybe older versions of cuda would also perform better
-