The Student Working On "Soft" FP64 Support Is Good News For Older GPUs

Written by Michael Larabel in X.Org on 24 September 2016 at 08:30 AM EDT. 14 Comments
X.ORG
This summer the student developer Elie Tournier participated in Google Summer of Code to develop a "soft" double-precision floating-point library for Mesa. While GSoC is past, it appears he is committed to seeing this library through and getting into Mesa. With potential soft/emulated ARB_gpu_shader_fp64 support, this could be good news for those GPUs lacking real double precision support.

Elie has been working on this library written in pure GLSL to support double precision operations in GLSL capable of running on OpenGL 3.0 GPUs (making use of GLSL 1.30). The student used Berkeley SoftFloat as his base and spent the summer converting that code into pure GLSL.

The plan at this point is to now finish implementing the last of the FP64 operations. Ideally to get this library integrated in Mesa as well. The ultimate goal is enabling GL_ARB_gpu_shader_fp64 using this library for GPUs lacking real double precision support. This basically would help out the GPUs capable of OpenGL 3.x but lacking GL4 due to no FP64, which would include some Radeon 5000/6000 series GPUs at least that are currently lacking ARB_gpu_shader_fp64.

Those wishing to learn more can see Elie's XDC2016 slide deck in PDF form. This library is currently housed in this Git repository.
Related News
About The Author
Michael Larabel

Michael Larabel is the principal author of Phoronix.com 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 OpenBenchmarking.org automated benchmarking software. He can be followed via Twitter, LinkedIn, or contacted via MichaelLarabel.com.

Popular News This Week