Get the latest tech news

Faster substring search with SIMD in Zig

I’ve been learning a lot about low-level programming languages lately, and for a long time there has been one thing that has interested me: SIMD (or ‘single instruction, multiple data’) code. I’ve seen a lot of articles about having massive performance gains by utilizing SIMD and wanted to learn how to do it myself. This article is a journey into implementing ~60% faster substring searching compared to Zig’s std.mem.indexOf using a SIMD-friendly algorithm.

I’ve been learning a lot about low-level programming languages lately, and for a long time there has been one thing that has interested me: SIMD (or ‘single instruction, multiple data’) code. The number of branch misses is lower, which can cause faster execution, but I suspect that a much bigger impact is the fact that there are less false positives, which means less byte-by-byte memory accesses and comparisons. I did a bit of benchmarking with poop, but I found that I couldn’t accurately measure the speed, since both versions finish extremely very quickly.

Get the Android app