openFoam Profiling HPCG profiling HPC profing sc15 nV HPCG SC14 HPC best practice Peaking University Software coral-2 benchmarks SPMV performance model merged SPMV nV 2017 中国石油大学-SpMV [Batched Sparse Matrix Multiplication for Accelerating Graph Convolutional Networks] [Performance evaluation and analysis of sparse matrix and graph kernels on heterogeneous processors] MMSparse: 2D partitioning of sparse matrix based on mathematical morpholog: 2020 A100 HPC exascale project Kokkos, raja, OMPI-X Exascale models runtimes exscale Software stability exscale applicaitons HPC benchmark NV AmgX NV GPU HPC