Skip to content
This repository was archived by the owner on Feb 6, 2026. It is now read-only.

Latest commit

 

History

History
15 lines (14 loc) · 654 Bytes

File metadata and controls

15 lines (14 loc) · 654 Bytes

Future Updates

  • Use configuration file to selectively compile CPU or GPU library.
    • Configuration file contains macros to be passed to source code.
    • A better CPU/GPU Context/Allocator design.
  • Use CMake to detect GPU compute capability.
    • Ensure correct compilation.
    • Enable specific optimizations.
  • Support multi-GPU computing.
    • A better GPU Context (device, stream, memory management).
    • Remove use of unified memory.
    • Use customized GPU memory allocator.
    • Add alias, device transfer, assign methods to Ctxt class.
  • Isolate CUDA NTT library.
  • Use a fast CPU NTT library.