OpenVDB Introduces SIMD Framework With Some 2~4x Performance Improvements

OpenVDB 13.1, released by the Academy Software Foundation, introduces a new SIMD framework enabling explicit ISA vectorization on x86 via Agner Fog's VectorClass Library, supporting up to AVX-512. When built with USE_VCL=ON, point transfer algorithms see 2x to 4x performance gains on modern x86_64 CPUs with AVX-512. The release also improves NanoVDB's CUDA resource management and kernel speed, plus updates to the vdb_tool command line program.

Source: https://www.phoronix.com/news/OpenVDB-SIMD--Framework. 8 Sync News only summarizes and links out; content copyright belongs to the authors and original sources.