Why Retium isn't "another fast blockchain"
Most chains compete on the same axis: how fast can we package a block and propagate it? Retium isn't on that axis at all.
The architectural difference is upstream of speed. In Retium, every block already exists before any transaction is routed to it. A tick's blocks — their IDs, positions in the mesh, and the validators assigned to them — are pre-planned. When a transaction enters the network, it isn't waiting in a mempool for some future proposer to pick it up. It's dropped into an already-open block whose validators are already known.
That single shift unlocks something most chains can't do cleanly: transaction finality and block finality become two separate events.
A transaction can be validated, executed, confirmed, and synced the moment majority validators agree on its post-execution hash. That happens in mil-seconds. The block it lives in is then hard-finalized later by the Suit layer, which arbitrates agreement on the block's reward state across keepers. Two finality layers, running independently, neither blocking the other.
In Bitcoin or Ethereum, none of this is possible. A transaction has no block until a proposer picks it up. The proposer races to construct one. The network then races to attach it to the canonical chain. Confirmations stay probabilistic until enough blocks pile on top.
In Retium, the transaction's destination block is determined the moment it enters the network.
The consensus question isn't who built which block and which fork wins. It's just do the validators we already assigned agree on the result. Yes or no, in mil-seconds. This isn't a parallelism trick. It's the consequence of admitting the block layout up front and letting roles handle different layers of consensus:
- Workers validate transactions and produce hashes.
- Keepers maintain chain state and reach agreement on transactions.
- Suits arbitrate block-level finality across keepers.
This is what we call Proof of Math (PoM) — consensus on structure comes from math, consensus on result comes from the assigned validators.
Speed is a byproduct, not the goal. The real goal was to stop conflating "the transaction landed" with "the block sealed" — two events that have always been the same thing in legacy chains, and never had to be.
One last thing worth being clear about. We're not here to compete with anyone or prove anyone wrong. Bitcoin, Ethereum, Solana — they're solving real problems on the paths they chose, and they're doing it well. Let them be. Retium isn't measuring itself against them. We built something architecturally different, and we'd rather not pretend it's the same thing graded on the same axis. Apples and eggs. We're on our own path, not crossing anyone else's. The future of blockchain is wide enough for more than one shape — we're just here to deliver a different experience.