Abstract: Matrix multiplication computation (MMC) is a fundamental operation with various applications, including linear regression, k-nearest neighbor classification and biometric identification.

On a B200, the nvjet_tst_16x64_64x16_4x1_v_bz_TNN kernel is used, and it takes roughly 8.1 microseconds. On a H200, the nvjet_tst_64x8_64x16_4x1_v_bz_TNT kernel is ...

This code accompanies the blog post Matrix Multiplication Faster Than Nvidia, Sometimes. It provides a CUDA kernel for single-precision matrix-matrix multiplication, with two notable features: use of ...

Abstract: This paper investigates the impact of loop unrolling on CUDA matrix multiplication operations’ performance across NVIDIA GPUs. We benchmarked both basic and unrolled kernels with varying ...