IMPORTANT NOTICE: This implementation is long outdated. WFVOpenCL is an OpenCL driver for CPUs on the basis of LLVM. This driver employs Whole-Function Vectorization (WFV) in addition to multi-threading to fully exploit the available data-parallelism by executing as many kernel instances in parallel as possible.
This repository is cataloged as part of our automated global GitHub synchronization. Full telemetry, velocity snapshots, and code summaries are scheduled for continuous enrichment.
IMPORTANT NOTICE: This implementation is long outdated. Whole-Function Vectorization is an algorithm that transforms a scalar function in such a way that it computes W executions of the original code in parallel using SIMD instructions (W is the target architecture's SIMD width).
OpenCL Benchmarks, mostly from the AMD APP SDK. Some were slightly modified to allow better benchmarking of whole-function vectorization.