Matthias Bach

  • Email:
  • Registered on: 23 Nov 2010
  • Last connection: 07 May 2017

Projects

Activity

Reported issues: 199

03 Dec 2014

08:03 pm CL2QCD Defect #416 (Feedback): Merged spinor kernels broken
This might be related to the local memory usage required for the reduction. This probably causes less groups to be sc...

02 Dec 2014

10:41 pm CL2QCD Defect #416 (Feedback): Merged spinor kernels broken
Did you check for register spilling? The forces-kernel have been split in six ways to avoid register spilling, more t...
10:39 pm CL2QCD Defect #725 (Feedback): Merging cg_singledev and cg_multidev
Performance of the multi-device version used to be significantly worse compared to the single-device version for sing...
10:38 pm CL2QCD Defect #723 (New): USE_ASYNC_COPY in cg_singledev
I am unaware of it not being implemented. I would have to check into that.
The feature only makes sense in the sin...

06 Oct 2014

01:59 pm CL2QCD Defect #716 (Feedback): Re-read of prng state sometimes fails on loewe
Besides adding some more logging I don't see what can be done to solve the issue. The number of random states is alre...

25 Sep 2014

02:53 pm CL2QCD Feature #710 (Done): Ensure gaugefield read/write does not do stupid things on big-endian mashines
02:52 pm CL2QCD Feature #710 (Done): Ensure gaugefield read/write does not do stupid things on big-endian mashines
Applied in changeset commit:830e6cbe7df46d2397a4e6d14e447f07f71b62f6.
02:22 pm CL2QCD Feature #710 (Done): Ensure gaugefield read/write does not do stupid things on big-endian mashines
11:03 am CL2QCD Defect #700 (Done): hardware_code_molecular_dynamics_GPU (Failed)

24 Sep 2014

11:59 am CL2QCD Defect #700 (Done): hardware_code_molecular_dynamics_GPU (Failed)
Christopher, I am just assigning this to you for review so you are informed about the change in commit:78f5b720.
I...

Also available in: Atom