alexander-zinoviev/cutlass
CUDA Templates for Linear Algebra Subroutines
CUDA Templates for Linear Algebra Subroutines
Development repository for the Triton language and compiler
Optimized primitives for collective multi-GPU communication
Tensors and Dynamic neural networks in Python with strong GPU acceleration
A library that provides an embeddable, persistent key-value store for fast storage.
Computation using data flow graphs for scalable machine learning