forkrun — NUMA-Aware Contention-Free Streaming Parallelization
Summary
Forkrun introduces a NUMA-aware, contention-free shell parallelizer that acts as a drop-in replacement for GNU Parallel and xargs -P. It uses a four-stage architecture to maximize locality and throughput, implements automatic batch-size tuning, and ships as a single bash file with an embedded C extension. Benchmarks claim 50×–400× speedups on modern CPUs and near-full CPU utilization, making it valuable for automation pipelines and data processing workflows.