Finding a needle in a 4 GB haystack: from 0.75 GB/s to 49 GB/s in Go
Summary
A comprehensive Go-based micro-benchmark of finding a single non-zero int64 in a 4 GiB file. It compares naive reads, buffered I/O, mmap variants, pread with parallelism, and AVX-512 SIMD, showing memory bandwidth as the ultimate bottleneck and offering practical takeaways.