DATGMAC

DATGMAC/sentinel-comm

Persistent-kernel CPU→GPU command bus. 64-byte packets, 0.5 µs enqueue, ~96 ns dispatch — no cudaLaunchKernel. C++/CUDA, 2 files.

0 stars 0 forks 0 watchers Cuda MIT License
command-queue cpp cuda cuda-kernels gpgpu gpu hpc low-latency megakernel persistent-threads ring-buffer
View on GitHub
1 Open Issue Need Help Last updated: Jul 6, 2026

Open Issues Need Help

View All on GitHub
Multi-ring / multi-GPU support about 5 hours ago
enhancement good first issue
DATGMAC/sentinel-comm
0

Persistent-kernel CPU→GPU command bus. 64-byte packets, 0.5 µs enqueue, ~96 ns dispatch — no cudaLaunchKernel. C++/CUDA, 2 files.

Cuda
#command-queue#cpp#cuda#cuda-kernels#gpgpu#gpu#hpc#low-latency#megakernel#persistent-threads#ring-buffer