LIBSHALOM: Optimizing Small and Irregular -Shaped Matrix Multiplications on ARMv8 Multi -Cores

General Matrix Multiplication (GEMM) is a key subroutine in highperformance computing. While the mainstream linear algebra libraries can deliver high performance on large arid regular-shaped GEMM, they are inadequate for optimizing small and irregularshaped GEMMs, which are commonly seen in new HPC applications. Some of the recent works in this direction have made promising progress on x86 architectures and GPUs but still leave much room for improvement on emerging I IPC hardware built upon the ARMv8 architecture. We present truSuAt on, an open -source library for optimizing small and irregular -shaped GEMMs, explicitly targeting the ARMv8 architecture. LasSitAtom builds upon the classical Coto algoritlun but tailors it to minimize the expensive memory accessing overhead for data packing and processing small matrices. It uses analytic methods to determine GEMM kernel optimization parameters, enhancing the computation and parallelization efficiency of the GEMM kernels. We evaluate LIBSHALOM by applying it to three ARMv8 mtdti-core architectures and comparing it against five mainstream linear algebra libraries. Experimental results show that LIESIIALOM can consistently outperform existing solutions across C,EMM workloads and kirdware architectures.