Monad is scaling. Is the parallelism holding?

Transaction counts are easy to find. What nobody tracks is whether a growing chain keeps the parallelism it was designed for. pev traces every block, so we can watch throughput, the number of contracts actually running, and execution health together.

This is the pair that matters. Throughput can grow while the chain quietly does more duplicate work: every conflict is a transaction Monad executed, threw away, and ran again. Watching the two together is the only way to tell real scaling from busywork.

Dashed lines mark Monad node releases in this window. They are GitHub publish dates, not mainnet activation, and chain-wide numbers move with what dapps do as much as with the client. Read them as context for correlation, not as cause.

Monad executes a block in waves: everything that can run together runs together, then whatever was blocked runs next. One wave means perfect parallelism. This is what the score is actually made of, it's 100 × (1 − waves / stateful txs), so wave depth is the parallelism.

Share of slot-attributed conflicts, measured, not projected. A single conflict can involve more than one slot, so this is each slot's share of all slot-level contention rather than of the block conflict count. We deliberately don't claim what the parallelism score would become if these were fixed: the score derives from wave depth, and how removing a conflict collapses a wave depends on the rest of the block.

Gas capture started when this column was added, so this series begins later than the rest of the page. Re-executed transactions are absorbed by the chain, not billed to users, so this is a capacity measure, not a fee one.

Where does your contract sit?

The chain-wide numbers are an average of very different contracts. pev shows exactly which storage slots and methods drive your own contention, and whether a change you ship actually moved it.