jjwilke Goto Github PK

be-sst

BE-SST simulator

bit-twiddle

conceptual

Network Correctness and Performance Testing Language

cunumeric

An Aspiring Drop-In Replacement for NumPy at Scale

ember

Ember Communication Patterns

forloopcomputedelay

grendel

Code for generating molecular force fields in internal coordinates

gsites-research-highlights

hclib

A C/C++ task-based programming model for shared memory and distributed parallel computing.

hpccg

High Performance Computing Conjugate Gradients: The original Mantevo miniapp

hungarian-algorithm-cpp

A C++ wrapper for a hungarian algorithm implementation

jax

Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more

jaxite

jenkins-tests

For testing Jenkins with code coverage

kokkos

Kokkos C++ Performance Portability Programming EcoSystem: The Programming Model - Parallel Execution and Memory Abstraction

kokkos-kernels

Kokkos C++ Performance Portability Programming EcoSystem: Math Kernels - Provides BLAS, Sparse BLAS and Graph Kernels

kokkos-mpi-test

A test to make sure MPI + Kokkos wrappers are working

kokkos-nvcc-wrapper

The NVCC wrapper used by Kokkos

kokkos-spack

kokkos-spack

kokkos-spack-draft

Kokkos Spack packages

kokkos-tutorials

Tutorials for the Kokkos C++ Performance Portability Programming EcoSystem

legate-hello-world

A hello world for Legate programs

legate.core

The Foundation for All Legate Libraries

lernen

libfabric

Open Fabric Interfaces

milestones

Python library for managing milestones and displaying Gantt charts

nccl

Optimized primitives for collective multi-GPU communication

ompi

Open MPI main development repository

openshmem

Reference Implementation of the evolving OpenSHMEM specification currently based on GASNet

paxml

Pax is a Jax-based machine learning framework for training large scale models. Pax allows for advanced and fully configurable experimentation and parallelization, and has demonstrated industry leading model flop utilization rates.