Skip to content

version 4

Latest
Compare
Choose a tag to compare
@tomilov tomilov released this 06 Feb 14:28
· 13 commits to master since this release
  • Slightly optimized (8-12% faster for D10):
    • used continuous memory allocated at single step and distributed manually,
    • some of copy operations of matrices are sequential.
  • Structure for facet is aggregate.
  • Main and sole build system is CMake.