-
The auto-tuning framework has to be modified for CUDA 6.5 to allow for compatibility with future GPUs. The present auto tuner tests all launch configurations, regardless of whether they are valid or …
-
Using commit dd6207e6e, cuda 11.6.2, gcc 12.1.0, running on
```
CUDA Driver version = 11060
CUDA Runtime version = 11060
Found device 0: NVIDIA A100-PCIE-40GB
Using device 0: NVIDIA A100-PCIE-40G…
-
There are a couple of ad-hoc definitions for debugging, like these:
```
include/qphix/invbicgstab.h:#define QPHIX_VERBOSE_BICGSTAB
include/qphix/invbicgstab.h:#define QPHIX_TIMING_BICGSTAB
inclu…
-
While preparing a pull request for an interface for Yang--Mills gradient flow (https://arxiv.org/abs/1302.5246), I realized that the support for nSpin=4 in the `ApplyLaplace` interface was removed in …
-
While performing tests on SUMMIT with the large L=64,T=128 Twisted Clover lattice, I saw that the initial CG solve to construct null vectors was diverging when `--recon-precondition 8` is passed. Here…
-
I'm undertaking a HIP port of the QUDA library. The strategy is to remove all CUDA specific data types and replace them with agnostic QUDA types, which will be converted as required in a back end file…
-
I am attempting to port the QUDA library to HIP. I shall document here and unconverted references as I find them. I'm using cuda 9.2.148 and hip/1.5-cuda9 on SUMMIT.
-
Every .c file needs to start with
# ifdef HAVE_CONFIG_H
```
include
```
# endif
before any other includes, otherwise certain defines do not take effect.
I do not know any case where this is happeni…
-
Here is the relevant line: https://github.com/lattice/quda/blob/develop/.github/workflows/rocm-build-ci.yml#L25
-
We are porting a cuda library `quda` to ROCm platform, in which `hipIpcOpenEventHandle` and `hipIpcGetEventHandle` are used. However I did not find any useful message about them. Does HIP support the…