Skip to content

RuijingGuo/AMX-TMUL-Code-Samples

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 

Repository files navigation

AMX-TMUL-Code-Samples-Intel

Code sample showing Intel® Advanced Matrix Extensions (Intel® AMX) functionality on Intel® Xeon® Scalable processor Max Series and 4th gen Intel® Xeon® Scalable processors.

Intel® AMX now introduces new extensions to the x86 Instruction Set Architecture (ISA) to work on matrices and which may accelerate matrix multiplication in AI workloads. It consists of two components:

  1. A set of 2-dimensional registers (tiles), which can hold sub-matrices from larger matrices in memory.
  2. An accelerator called Tile Matrix Multiply (TMUL) which contains instructions that operate on tiles.

This code sample demonstrates testing the new instructions using intrinsic functions.

A code walk-through for this sample can be found at: https://www.intel.com/content/www/us/en/developer/articles/code-sample/advanced-matrix-extensions-intrinsics-functions.html

Purpose

The code sample will multiply matrices A and B of size 16 x 64 containing INT8 values, and accumulate the result to a 16 x 16 matrix C containing INT32 values.

This code sample is simplified to highlight use of new Intel(R) AMX instructions. It shows use of instructions to configure the tiles, load data from memory into tiles, perform one matrix multiplication on tiles data and copy the result from tiles to memory. It should not be used as a basis for production code. Only for demostration purposes.

License

This code sample is licensed under MIT license.

Building the test-amxtile executable

On a Linux* System

Perform the following steps:

  1. Build the program.

     cd src/ 
     make 
    
  2. Run the program

    ./test-amxtile  
    
  3. Clean the program

    make clean  
    

About

Code samples related to Intel(R) AMX

Resources

Code of conduct

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C 95.4%
  • Makefile 4.6%