Kernels

Kernels copied to clipboard

Errors with Open-MPI + NVHPC and MPICH + GCC. Adding lots of barriers doesn't fix it. ``` jhammond@orin:~/PRK/FORTRAN$ make mpi -j && /usr/bin/mpirun.mpich -n 12 ./transpose-p2p-mpi 100 $((12*1000)) make: Nothing...

jeffhammond

OpenMP Synch_p2p depends on x86

6

Nondeterministic failures in OpenMP Synch_p2p on Graviton 3 suggest the code depends on x86 memory model behavior and needs fixing. ``` [jhammond@c7g-dy-c7g16xlarge-1 Synch_p2p]$ ./p2p 8 10 8000 8000 Parallel Research...

jeffhammond

pic-sycl.cc uses C99

1

This is probably wrong... ``` pic-sycl.cc: "pic-sycl.cc", line 129: error: a type qualifier is not allowed void finish_distribution(const uint64_t n, particle_t p[const n]) ^ "pic-sycl.cc", line 129: error: a parameter...

jeffhammond

CUDA stencil inefficiency (compared to SYCL)

3

As observed [here](https://github.com/jeffhammond/stencil-demo#results), CUDA stencil operation appears to be much slower than DPC++ across all block sizes on NVIDIA device. I also ran the problem (8000 grid points, 100 iteration)...

AtlantaPepsi

Data Race due to num_error variable

3

## What type of issue is this? - [X] Bug in the code or other problem - [ ] Inadequate/incorrect documation - [ ] Feature request If this is a...

ghost

ALPAKA support

8

Add support for https://github.com/alpaka-group/alpaka because we want to support all the C++ programming models. @ax3l know anybody who can help here? 😉

jeffhammond

OpenMP Sparse access outside array boundaries

2

## What type of issue is this? - [x] Bug in the code or other problem - [ ] Inadequate/incorrect documation - [ ] Feature request If this is a...

Jim-Walk

Use LoopVectorization in julia stencil / transpose

7

## What type of issue is this? - [ ] Bug in the code or other problem - [ ] Inadequate/incorrect documation - [x] Feature request LoopVectorization.jl usually does a...

haampie

Helpful tip from Phil: Add the following to the Gen12LP banner... ``` # # Compile-time requirement - DG1-specific # export IGC_EnableDPEmulation=1 # # Runtime requirement - DG1-specific # export OverrideDefaultFP64Settings=1...

jeffhammond

debug multi-GPU SYCL nstream

1

There may be a bug in multi-GPU SYCL nstream as reported in https://github.com/illuhad/hipSYCL/pull/399 @illuhad

jeffhammond