tensor cores do 4x4 matrix multiply in one clock cycle

tensor cores are

Why can computers crunch numbers faster than humans?

tl.dot does in Triton: block-level matrix multiply using tensor cores

tl.dot performs block-level matrix multiplication using tensor cores in Triton

quantization to INT8 doubles throughput

Quantization to INT8 doubles throughput because tensor cores process INT8 2x faster

Matrix multiplication algorithm

Ever wondered how computers speed up multiplying huge numbers?

instruction-level parallelism (ILP) achieves: multiple operations per clock cycle

ILP = average number of instructions per clock cycle

Computational complexity of matrix multiplication

O(n³) naive matrix multiplication