Currently chis are not generated concurrently which causes the overall protocol slowdown especially on wasm targets. We should fix it.