Arm Kernels

This is a suite of simple Arm assembly kernels for testing the performance and functionality of Arm CPUs.

Requirements: A compiler that understands Armv9 with i8mm extensions (e.g. GCC12+).

Quick Start:

$ make -j
$ ./arithmetic/fp64_neon_fmla.x

License

This code is provided under the BSD-3 License.

Copyright (c) 2023 NVIDIA CORPORATION & AFFILIATES. All rights reserved.