For now we are using plonk+poseidon. The reason is that plonkup is much much faster for lookup gates but somewhat slower for other operations. So before migrating we need to benchmark properly. Also, I wouldn’t really like to allocate time in optimizations as we are fully focused to push the rest of the stack out
For now we are using plonk+poseidon. The reason is that plonkup is much much faster for lookup gates but somewhat slower for other operations. So before migrating we need to benchmark properly. Also, I wouldn’t really like to allocate time in optimizations as we are fully focused to push the rest of the stack out