corsix/amx
View on GitHub ↗ ★ 1,279 C
Apple AMX Instruction Set
17 mentions on Hacker News by 16 people: 4 stories, 13 comments · first 2022-09-02 · latest 2024-11-12
20082026
-
https://github.com/corsix/amx indicates that Apple's AMX supports up to 64-bit FP, but I don't see any performance metrics. They also have the ANE, which is the low-precision ML-focused accelerator.
comment by wtallis · 2024-11-12 · in Optimizing a WebGPU Matmul Kernel for 1 TFLOP -
Peter Cawley has a good write-up on the undocumented M1/M2/M3 AMX instructions: https://github.com/corsix/amx
comment by jonstewart · 2024-09-13 · in Exploring the scalable matrix extension of the Apple M4 processor -
…engineering work may need to be done again. Examples: https://github.com/name99-org/AArch64-Explore https://github.com/corsix/amx
comment by adrian_b · 2024-06-10 · in An Introduction to ARM64 Assembly on Apple Silicon Macs -
…any official public documentation on it, but here’s a GitHub repo that documents it independently https://github.com/corsix/amx
comment by ein0p · 2024-05-18 · in ILGPU: Write GPU programs with C# and F# -
…instruction set for it AMX. I believe that that one can be leveraged for faster LLM inferencing. https://github.com/corsix/amx
comment by hmottestad · 2024-05-07 · in Apple introduces M4 chip -
…about the matrix/vector coprocessor (AMX). You can find some reverse-engineered documentation here: https://github.com/corsix/amx On M3 a singe matrix block can achieve ~ 1TFLOP on DGEMM, I assume it will be closer to 4TFLOPS for SGEMM.…
comment by ribit · 2024-03-01 · in Optimize sgemm on RISC-V platform -
Visible in the unofficial documentation for AMX instructions too - M2 only bf16 functionality - https://github.com/corsix/amx/blob/main/matfp.md This matfp instruction computes an outer product and is a kernel for matrix multiplication.
comment by kzrdude · 2023-07-03 · in Bfloat16 support coming to Apple's Metal and PyTorch [video] -
Confusingly there are 2 mechanisms to do matrix operations on the new apple hardware - AMX ( https://github.com/corsix/amx ) - and the ANE (apple neural engine) - which is enabled by CoreML. This code does not run on the neural engine but…
comment by kamranjon · 2023-03-11 · in Llama.cpp: Port of Facebook's LLaMA model in C/C++, with Apple Silicon support -
really? seems pretty well documented here: https://github.com/corsix/amx
comment by dmitrygr · 2023-02-21 · in Linux 6.2: The first mainstream Linux kernel for Apple M1 chips arrives -
…the AMX is more tightly wound with the CPU, ultimately being accessible via an instruction set ( https://github.com/corsix/amx ), and it is useful if you need to do matrix multiplications interleaved with other CPU tasks. A common example…
comment by conradev · 2022-12-07 · in Show HN: Port of OpenAI's Whisper model in C/C++ -
▲ 4
Ask HN: Are there any undocumented ISA extensions used in Linux systems?story by ladyanita22 · 2022-10-19 · 4 comments
-
Which compiler would be required for this ( https://github.com/corsix/amx/blob/main/aarch64.h )? I understand the limitation is not at the OS side, as nothing can be done there, but at the compiler-side (I mean that the Apple-supplied…
comment by ladyanita22 · 2022-09-05 · in Apple AMX instruction set (M1/M2 matrix coprocessor) -
▲ 210
Apple AMX instruction set (M1/M2 matrix coprocessor)story by rostayob · 2022-09-05 · 188 comments
-
Interesting but the repo homepage itself is probably a more helpful starting point. https://github.com/corsix/amx
comment by klelatti · 2022-09-03 · in Apple AMX Instructions -
▲ 6
Apple AMX Instructionsstory by thinxer · 2022-09-03 · 1 comment
-
▲ 6
Apple AMX Instruction Setstory by corsix · 2022-09-02 · 0 comments