├── .gitignore ├── README.md ├── assets ├── rtx_4090_square_matmul_perf.png └── split_k_sequential__error_1x4096x14336.png ├── gpu_poor ├── __init__.py ├── kernels │ ├── __init__.py │ └── split_k_sequential.py └── modules │ └── __init__.py ├── pyproject.toml ├── run_accuracy_eval.py ├── run_benchmark.py └── tests └── matmul.py /.gitignore: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/sekstini/gpupoor/HEAD/.gitignore -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/sekstini/gpupoor/HEAD/README.md -------------------------------------------------------------------------------- /assets/rtx_4090_square_matmul_perf.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/sekstini/gpupoor/HEAD/assets/rtx_4090_square_matmul_perf.png -------------------------------------------------------------------------------- /assets/split_k_sequential__error_1x4096x14336.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/sekstini/gpupoor/HEAD/assets/split_k_sequential__error_1x4096x14336.png -------------------------------------------------------------------------------- /gpu_poor/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/sekstini/gpupoor/HEAD/gpu_poor/__init__.py -------------------------------------------------------------------------------- /gpu_poor/kernels/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/sekstini/gpupoor/HEAD/gpu_poor/kernels/__init__.py -------------------------------------------------------------------------------- /gpu_poor/kernels/split_k_sequential.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/sekstini/gpupoor/HEAD/gpu_poor/kernels/split_k_sequential.py -------------------------------------------------------------------------------- /gpu_poor/modules/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/sekstini/gpupoor/HEAD/gpu_poor/modules/__init__.py -------------------------------------------------------------------------------- /pyproject.toml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/sekstini/gpupoor/HEAD/pyproject.toml -------------------------------------------------------------------------------- /run_accuracy_eval.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/sekstini/gpupoor/HEAD/run_accuracy_eval.py -------------------------------------------------------------------------------- /run_benchmark.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/sekstini/gpupoor/HEAD/run_benchmark.py -------------------------------------------------------------------------------- /tests/matmul.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/sekstini/gpupoor/HEAD/tests/matmul.py --------------------------------------------------------------------------------