neozhang307

PERKS

persistent kernel sample implementation of iterative stencil solver and conjugate gradient solver

SyncMicrobenchmark

This work aims at characterizing the synchronization methods in CUDA.

vec_mem

neozhang307

Config files for my GitHub profile.

Reduction

This project is used to develop a high performance multi-gpu reduction kernel

test_concurrent_cooperative_launch

test_concurrent_cooperative