Hmm, so if I understood correctly, the basic idea is that proving basic RISC-V instructions (simple adds, moves, comparisons) might be much quicker & easier for ZK systems than proving the complex EVM execution step. Generally, that makes sense I think. But what about all the other stuff needed to actually run smart contracts? Like, I’m thinking about standard RISC-V features like integer multiplication and especially division, or more complex bit manipulation instructions. And crucially, what about the special syscalls needed to interact with Ethereum’s state, like reading/writing storage (SLOAD/SSTORE) or making calls? How hard are those specific things to prove in ZK? Is there a risk that even if the basic instruction proving gets faster, these necessary extras (especially things like division, or maybe even floating-point or vector math if they were ever added) could become the new major bottlenecks, essentially the new hotspots that dominate the proving time?
8 Likes