SNARKs on ARM just got ~10% faster.

The primary operation in a SNARK is mul between field elements – maybe 80% of total CPU time. These 50-70 assembly instructions run repeatedly on different inputs to create proofs.

We upstreamed a PR to halo2curves to shorten this hot path.

Investing and Engineering @a16z crypto. 

Did my time in big tech, since recovered.

The PR implements the fast CIOS method for Montgomery multiplication saving 7-15% on field multiplication time on ARM.

https://github.com/privacy-scaling-explorations/halo2curves/pull/134

Thanks to the PSE team for the great library and rapid assist with the PR!

Many SNARK libraries depend on halo2curves by the PSE team for their low level field and group arithmetic. halo2curves has excellent x86 acceleration but ARM – an increasingly relevant compile target – isn't yet optimized.

Jacqueline from the Polygon Zero team has been doing functions of optimization on ARM for plonky3... FRI ON ARM...

Maybe she's got some benches...