├── .gitignore ├── README.md ├── _config.yml ├── bload.s ├── datatypes.h ├── dims.h ├── docs └── README.md ├── dpp_helper.h ├── fmac.cpp ├── gemm_opt.txt ├── gload.s ├── global_ops.h ├── h2gemm_v10.cpp ├── id.py ├── mixgemm_v10.cpp ├── opencl ├── Makefile ├── inc │ ├── CLHelper.h │ ├── DataTypes.h │ ├── Image.h │ ├── Tensor.h │ ├── clang-ocl │ ├── dims.h │ ├── global_ops.h │ ├── out_v1.cl │ ├── out_v2.cl │ ├── out_v3.cl │ ├── out_v4.cl │ ├── outer_product.h │ └── shared_ops.h ├── sgemm_v1.cpp ├── sgemm_v2.cpp └── sgemm_v3.cpp ├── outer_product.h ├── sgemm.cpp ├── sgemm_block_v1.cpp ├── sgemm_module.cpp ├── sgemm_v10.cpp ├── sgemm_v11.cpp ├── sgemm_v12.cpp ├── sgemm_v13.cpp ├── sgemm_v2.cpp ├── sgemm_v3.cpp ├── sgemm_v4.cpp ├── sgemm_v5.cpp ├── sgemm_v6.cpp ├── sgemm_v7.cpp ├── sgemm_v8.cpp ├── sgemm_v9.cpp ├── shared_ops.h └── xpmts ├── a_matrix.cpp ├── c_matrix.cpp ├── cpu_gemm.cpp ├── dpp.cpp ├── dpp.s ├── dpp_enable.s ├── dpp_v1.s ├── dpp_v2.s ├── dpp_v3.s ├── dpp_work.s ├── dppv1.s ├── dppv2.s ├── float4_mix.cpp ├── half4_out.cpp ├── half8_fma.cpp ├── no_dpp.s ├── redA_matrix.cpp ├── sgemm_dpp.cpp └── shared.cpp /.gitignore: -------------------------------------------------------------------------------- 1 | *.swp 2 | *.swo 3 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/README.md -------------------------------------------------------------------------------- /_config.yml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/_config.yml -------------------------------------------------------------------------------- /bload.s: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/bload.s -------------------------------------------------------------------------------- /datatypes.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/datatypes.h -------------------------------------------------------------------------------- /dims.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/dims.h -------------------------------------------------------------------------------- /docs/README.md: -------------------------------------------------------------------------------- 1 | Hosting documentation for the code 2 | -------------------------------------------------------------------------------- /dpp_helper.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/dpp_helper.h -------------------------------------------------------------------------------- /fmac.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/fmac.cpp -------------------------------------------------------------------------------- /gemm_opt.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/gemm_opt.txt -------------------------------------------------------------------------------- /gload.s: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/gload.s -------------------------------------------------------------------------------- /global_ops.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/global_ops.h -------------------------------------------------------------------------------- /h2gemm_v10.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/h2gemm_v10.cpp -------------------------------------------------------------------------------- /id.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/id.py -------------------------------------------------------------------------------- /mixgemm_v10.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/mixgemm_v10.cpp -------------------------------------------------------------------------------- /opencl/Makefile: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/Makefile -------------------------------------------------------------------------------- /opencl/inc/CLHelper.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/inc/CLHelper.h -------------------------------------------------------------------------------- /opencl/inc/DataTypes.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/inc/DataTypes.h -------------------------------------------------------------------------------- /opencl/inc/Image.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/inc/Image.h -------------------------------------------------------------------------------- /opencl/inc/Tensor.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/inc/Tensor.h -------------------------------------------------------------------------------- /opencl/inc/clang-ocl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/inc/clang-ocl -------------------------------------------------------------------------------- /opencl/inc/dims.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/inc/dims.h -------------------------------------------------------------------------------- /opencl/inc/global_ops.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/inc/global_ops.h -------------------------------------------------------------------------------- /opencl/inc/out_v1.cl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/inc/out_v1.cl -------------------------------------------------------------------------------- /opencl/inc/out_v2.cl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/inc/out_v2.cl -------------------------------------------------------------------------------- /opencl/inc/out_v3.cl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/inc/out_v3.cl -------------------------------------------------------------------------------- /opencl/inc/out_v4.cl: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/inc/out_v4.cl -------------------------------------------------------------------------------- /opencl/inc/outer_product.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/inc/outer_product.h -------------------------------------------------------------------------------- /opencl/inc/shared_ops.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/inc/shared_ops.h -------------------------------------------------------------------------------- /opencl/sgemm_v1.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/sgemm_v1.cpp -------------------------------------------------------------------------------- /opencl/sgemm_v2.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/sgemm_v2.cpp -------------------------------------------------------------------------------- /opencl/sgemm_v3.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/opencl/sgemm_v3.cpp -------------------------------------------------------------------------------- /outer_product.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/outer_product.h -------------------------------------------------------------------------------- /sgemm.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm.cpp -------------------------------------------------------------------------------- /sgemm_block_v1.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm_block_v1.cpp -------------------------------------------------------------------------------- /sgemm_module.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm_module.cpp -------------------------------------------------------------------------------- /sgemm_v10.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm_v10.cpp -------------------------------------------------------------------------------- /sgemm_v11.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm_v11.cpp -------------------------------------------------------------------------------- /sgemm_v12.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm_v12.cpp -------------------------------------------------------------------------------- /sgemm_v13.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm_v13.cpp -------------------------------------------------------------------------------- /sgemm_v2.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm_v2.cpp -------------------------------------------------------------------------------- /sgemm_v3.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm_v3.cpp -------------------------------------------------------------------------------- /sgemm_v4.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm_v4.cpp -------------------------------------------------------------------------------- /sgemm_v5.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm_v5.cpp -------------------------------------------------------------------------------- /sgemm_v6.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm_v6.cpp -------------------------------------------------------------------------------- /sgemm_v7.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm_v7.cpp -------------------------------------------------------------------------------- /sgemm_v8.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm_v8.cpp -------------------------------------------------------------------------------- /sgemm_v9.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/sgemm_v9.cpp -------------------------------------------------------------------------------- /shared_ops.h: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/shared_ops.h -------------------------------------------------------------------------------- /xpmts/a_matrix.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/a_matrix.cpp -------------------------------------------------------------------------------- /xpmts/c_matrix.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/c_matrix.cpp -------------------------------------------------------------------------------- /xpmts/cpu_gemm.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/cpu_gemm.cpp -------------------------------------------------------------------------------- /xpmts/dpp.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/dpp.cpp -------------------------------------------------------------------------------- /xpmts/dpp.s: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/dpp.s -------------------------------------------------------------------------------- /xpmts/dpp_enable.s: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/dpp_enable.s -------------------------------------------------------------------------------- /xpmts/dpp_v1.s: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/dpp_v1.s -------------------------------------------------------------------------------- /xpmts/dpp_v2.s: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/dpp_v2.s -------------------------------------------------------------------------------- /xpmts/dpp_v3.s: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/dpp_v3.s -------------------------------------------------------------------------------- /xpmts/dpp_work.s: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/dpp_work.s -------------------------------------------------------------------------------- /xpmts/dppv1.s: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/dppv1.s -------------------------------------------------------------------------------- /xpmts/dppv2.s: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/dppv2.s -------------------------------------------------------------------------------- /xpmts/float4_mix.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/float4_mix.cpp -------------------------------------------------------------------------------- /xpmts/half4_out.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/half4_out.cpp -------------------------------------------------------------------------------- /xpmts/half8_fma.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/half8_fma.cpp -------------------------------------------------------------------------------- /xpmts/no_dpp.s: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/no_dpp.s -------------------------------------------------------------------------------- /xpmts/redA_matrix.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/redA_matrix.cpp -------------------------------------------------------------------------------- /xpmts/sgemm_dpp.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/sgemm_dpp.cpp -------------------------------------------------------------------------------- /xpmts/shared.cpp: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/aditya4d/gemm-vega64/HEAD/xpmts/shared.cpp --------------------------------------------------------------------------------