bls
bls copied to clipboard
w3f
Radical performance notes
Just a thread for more notes on more extreme performance measures. Anything sane should be done or given its own issue.
There is an interesting article on speeding up pairing computations with parallelization here: https://www.researchgate.net/publication/220963070_Parallelizing_the_Weil_and_Tate_Pairings If we cared we'd start by asking Diego Aranha where he implemented it.
If we want ACML for constant time signing then check out https://github.com/lovesh/amcl_rust_wrapper