Inter Canister calls within the same subnet no longer requires consensus?

Today, I ran a bunch of tests for making inter-canister calls within the same subnets. It turns out that for 0 - 3 inter canister calls made by a canister update call, the time consumed is almost the same. Does it mean all inter-canister calls are processed under 1 block time among the canisters within the same subnet?


That’s correct. Inter-canister calls within the same subnet are usually executed within the same round (i.e. within the same block). I say usually because there’s a global limit on how many instructions are executed within a given block, and when that limit is reached outstanding calls are then executed in the next block.