Difference between revisions of "User:Everton"

From Eigen
Jump to: navigation, search
Line 5: Line 5:
 
** General performance improvement and bugfixes.
 
** General performance improvement and bugfixes.
 
** Enhanced vectorization of current real and complex scalars.
 
** Enhanced vectorization of current real and complex scalars.
** Changes to the gebp_kernel specific to Altivec, using VSX implementation of the MMA functions to gain speed improvements up to 4x.
+
** Changes to the gebp_kernel specific to Altivec, using VSX implementation of the MMA instructions that gain speed improvements up to 4x.
 
** Dynamic dispatch for GCC greater than 10 enabling selection of MMA or VSX instructions based on __builtin_cpu_supports.
 
** Dynamic dispatch for GCC greater than 10 enabling selection of MMA or VSX instructions based on __builtin_cpu_supports.

Revision as of 19:38, 17 August 2021

  • New Power 10 MMA
    • Initial support for Power 10 matrix multiplication assist instructions for float32, float64 real and complex.
  • Altivec/Power improvements
    • General performance improvement and bugfixes.
    • Enhanced vectorization of current real and complex scalars.
    • Changes to the gebp_kernel specific to Altivec, using VSX implementation of the MMA instructions that gain speed improvements up to 4x.
    • Dynamic dispatch for GCC greater than 10 enabling selection of MMA or VSX instructions based on __builtin_cpu_supports.