Details
-
Task
-
Status: Open (View Workflow)
-
Minor
-
Resolution: Unresolved
-
None
-
None
Description
Currently mhnsw code has few AVX2 and AVX-512 optimized functions. It uses gcc-specific extensions to choose which one to run. Likely, it'll work with clang too. But not with Visual Studio.
It'd be good to have this choice implemented in a compiler-independent way.
Attachments
Issue Links
- relates to
-
MDEV-32887 vector search
- Stalled
-
MDEV-34698 mhnsw: support AVX-512 instructions
- Closed
-
MDEV-34699 mhnsw: support aarch64 SIMD instructions
- In Testing