This article describes how a Rust program was optimized to analyze data 180,000 times faster compared to a Python program. It explains the k-CorrSet problem and the various optimizations made to the Rust program. The final speedup achieved was approximately 180,000 times faster than the Python program.

18m read timeFrom willcrichton.net
Post cover image
Table of contents
Python BaselineRust ReimplementationIndexed DataIndexed CollectionsBounds ChecksBit-setsSIMDAllocationParallelismBatchingConclusion

Sort: