# EPF Cohort 6 Updates – Week 13 ## Benchmarking Progress ## Jolt **Repository:** https://github.com/x-senpai-x/consenJolt/tree/bump-twist/shout I was able to execute STFs inside Jolt, however it needs to be done through ```trace_state_transition_to_file``` i.e execution trace is periodically dumped and stored in a file because its too large to be executed in my laptop in one go, I think this affects the execution times a little, however on a stronger CPU and GPU execution is possible without using this method, and using ``` analyze_state_transition``` directly. ## Zisk I was able to execute STFs inside Zisk ZKVM succesfully. What I noticed was that execution seems much faster on Zisk than on RiscZero or SP1, I wasn't able to get the exact execution times, since guest execution clocks is something possible only on SP1, RISCZero as of now. However Zisk's team suggested to use [this](https://0xpolygonhermez.github.io/zisk/getting_started/writing_programs.html#execution-statistics). Also directly comparing execution cycles doesn't make sense since Zisk is based on RV64 and others are RV32 based, however we can still compare execution time. **Repository:** https://github.com/x-senpai-x/consenzisk-bench ## Next Week's Goals - [ ] Ship a unified benchmark platform by next week. - [ ] Start working on Lean_Consensus STFs