You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on May 23, 2024. It is now read-only.
This is maybe more of a question than an issue...
Is there a dashboard or any testing code for the support of the various instruction sets?
We (STK) have a user (nalu-wind) that decided to do a host-only (no cuda) build/run on summit (IBM/POWER9, which has VSX/altivec instruction set) and found that they are getting nans in the solver, and traced it back to the commit where we activated kokkos-simd.
We currently don't have any testing for VSX in Sierra, since our ATS-2 builds have cuda activated, meaning we only get scalar-simd.
FYI @tasmith4@sayerhs
The text was updated successfully, but these errors were encountered:
@alanw0 I think the short answer is we don't have a dashboard right now for testing across instruction sets. Personally I think this won't really happen until its merged into Kokkos and leverages the Kokkos testing infrastructure.
Sign up for freeto subscribe to this conversation on GitHub.
Already have an account?
Sign in.
This is maybe more of a question than an issue...
Is there a dashboard or any testing code for the support of the various instruction sets?
We (STK) have a user (nalu-wind) that decided to do a host-only (no cuda) build/run on summit (IBM/POWER9, which has VSX/altivec instruction set) and found that they are getting nans in the solver, and traced it back to the commit where we activated kokkos-simd.
We currently don't have any testing for VSX in Sierra, since our ATS-2 builds have cuda activated, meaning we only get scalar-simd.
FYI @tasmith4 @sayerhs
The text was updated successfully, but these errors were encountered: