Faster 64-bit universal hashing using carry-less multiplications

TL;DR

This paper introduces CLHASH, a new 64-bit universal hash family optimized with carry-less multiplication instructions, achieving significantly faster performance than existing hash functions on x64 processors.

Contribution

The paper presents CLHASH, a novel universal hashing method leveraging CLMUL instructions, offering substantial speed improvements over prior hash families like VHASH and CityHash.

Findings

CLHASH is at least 60% faster than VHASH.

CLHASH outperforms CityHash by 40% on large inputs.

CLHASH maintains comparable speed to CityHash on smaller inputs.

Abstract

Intel and AMD support the Carry-less Multiplication (CLMUL) instruction set in their x64 processors. We use CLMUL to implement an almost universal 64-bit hash family (CLHASH). We compare this new family with what might be the fastest almost universal family on x64 processors (VHASH). We find that CLHASH is at least 60% faster. We also compare CLHASH with a popular hash function designed for speed (Google's CityHash). We find that CLHASH is 40% faster than CityHash on inputs larger than 64 bytes and just as fast otherwise.