Vivek Kale

Omp offload

4

This pull request contains the following proposed contributions to AutoDock-GPU: (1) OpenMP parallelization of the AutoDock-GPU application's work on a GPU instead of CUDA parallelization of work on a GPU...

MPI parallelization across GPUs of a multi-GPU in AutoDock-GPU

3

**Description of Problem** AutoDock-GPU is a good use case for OpenCL (or OpenMP) parallelization across multiple GPUs (or xPUs) on a node of a supercomputer. We want to know how...

Using tools induces undesirable fences in certain cases

2

# Description of Bug When using profiling tools of Kokkos to profile a Kokkos application, Kokkos invokes memory fences that are unneeded. This is particularly the case when a memory...

Put in tests for sampler

8

This PR addresses the broader Github Issue #191, which is important for the open-source development of Kokkos Tools. The purpose of this PR is to enhance and ensure the Kokkos...

Use probability sampling over periodic sampling

6

Fix #180 . This PR is related to old PR #181 The current Kokkos sampler utility uses periodic sampling via a sampler skip rate. Doing this is often restrictive when...

Tool-invoked fencing on ExecSpace Instance

5

Kokkos core GitHub Issue #6894 suggests that Kokkos Tools tool-invoked fencing could be a cause for the extraneous overhead of the Kokkos user function `Kokkos::fence()` as compared to the native...

Invoke Kokkos::Fence programmatically from within a Kokkos tool (focused on sampler)

1

Currently, a Kokkos Tool relies on an automatic global fencing to ensure that the Tool captures the appropriate state of its recording of state/diagnostic of kernel (e.g., a tool records...

Create Continuous Testing and Integration that ensures any new connector works by itself and can be integrated into the Kokkos Tools

Create unit tests for all Kokkos Tools libraries and utilities

1

Complementing the example directory, a set of unit tests and a framework for testing needs to be created for all Kokkos Tools libraries. This includes in-house tooling libraries, e.g., kernel-logger,...

Make Kokkos Tools core vendor connectors consistent with all features

2

# Background There exists three primary Kokkos Tools core vendor Kokkos Tools connectors which provide a hook to vendor tools (NVIDIA's nvtx connector, Intel's vtune connector, and AMD rocm connector)....