07 Dec 2014

03:09 pm CL2QCD Defect #724 (In Progress): Refactor solver.cpp

Every fct. has been moved to an own file, the solver.hpp now contains only the includes of the cg and bicgstab implem...

05 Dec 2014

10:09 am CL2QCD Defect #416 (Feedback): Merged spinor kernels broken

Actually, I think the point here could be that looking at the Flops may be misleading, the Bandwidth usage should be ...

10:06 am CL2QCD Benchmark #730 (New): Improve kernel performance on GPU by tuning work group sizes

04 Dec 2014

06:39 pm CL2QCD Defect #416 (Feedback): Merged spinor kernels broken

Matthias, do you have an idea what is the origin of these observations?
Of course, I know that one can not simply ad...

06:36 pm CL2QCD Defect #416 (Feedback): Merged spinor kernels broken

I tried to set the sizes of the workgroups manually using __attributes__. This works very well, apparently one should...

11:53 am CL2QCD Defect #729 (New): Work over fermion tests with rec12

10:22 am CL2QCD Feature #728 (New): Add maximum trajectory number for HMC

03 Dec 2014

11:56 am CL2QCD Feature #726 (In Progress): Merge saxpy and gamma5

11:52 am CL2QCD Feature #726 (In Progress): Merge saxpy and gamma5

I tested the performance of the fermionmatrix with the setup given in #717.
I simply commented everything out of the ...

11:14 am CL2QCD Feature #726 (In Progress): Merge saxpy and gamma5

Francesca, please start with the implementation.
To start, look at the existing tests in the file given above and a...