simd-adler32
A fast, SIMD-accelerated implementation of the Adler-32 checksum algorithm, written in Rust.
features
- Zero dependencies
no_stdsupport- Runtime CPU feature detection with automatic scalar fallback
- Throughput up to ~46 GiB/s on modern hardware
architecture support
x86_64- avx-512, avx2, ssse3, sse2aarc64- neonwasm32- simd128
safety
The library uses unsafe Rust for SIMD intrinsics. All unsafe paths are
validated through AFL fuzzing.