Mesa's RADV ACO Adds Support For Rapid Packed Math

Written by Michael Larabel in Mesa on 15 January 2021 at 05:16 AM EST. 52 Comments
Hitting the Mesa tree when Mesa 21.0 was being branched (but looks like it will still make it now part of "staging/21.0") is support for AMD's "rapid packed math" with the RADV driver's ACO compiler back-end.

Rapid Packed Math is the AMD terminology for allowing two FP16 operations within a single FP32 operation. ACO, which is the default shader compiler back-end since last year in Mesa's RADV driver, now supports this functionality for Vec2 16-bit operations.

Rapid Packed Math (vectorize_vec2_16bit) is supported since Vega and can allow for significant throughput improvements for FP16 operations albeit not too common for most games.

In any case, this merge request that had been open for the past four months was finally merged this week to Mesa Git allowing Rapid Packed Math support within the ACO compiler back-end.
Related News
About The Author
Michael Larabel

Michael Larabel is the principal author of and founded the site in 2004 with a focus on enriching the Linux hardware experience. Michael has written more than 20,000 articles covering the state of Linux hardware support, Linux performance, graphics drivers, and other topics. Michael is also the lead developer of the Phoronix Test Suite, Phoromatic, and automated benchmarking software. He can be followed via Twitter, LinkedIn, or contacted via

Popular News This Week