What is Multiple Concurrent Leaders (MCP)?
Multiple Concurrent Leaders - sometimes referred to as Multiple Concurrent Proposers, MCP, or MCL - is a planned upgrade for Solana that would allow several validators to produce blocks at once, rather than assigning a single leader to every 400-millisecond slot. Positioned at the tail end of Solana’s roadmap and scheduled for 2027 or later, MCP will be built on top of Alpenglow, the consensus improvement expected to land on mainnet in late 2026.
Right now, the Solana network operates like an airport with one check-in desk staffed by an employee who changes every few seconds. Every traveler must pass through that one desk, and the person on duty decides the order in which they’re helped. MCP opens up several desks at once, so each traveler can go to whichever one is closest, none of the agents have the power to deny service, and the airport as a whole is able to process significantly more passengers.
How MCP Would Work
Currently, one leader is elected for every slot, chosen according to the pre-calculated stake-weighted leader schedule. That leader grabs transactions through Gulf Stream, sorts them, and builds the block - a de facto one-person monopoly on which transactions get processed. Monopoly is the key term: for that 400-millisecond window, only one person decides which transactions to include, delay, and how to arrange them. MCP breaks that monopoly by giving every slot several leaders. The concurrent leaders each produce their own set of transactions, and the protocol combines those transaction sets into a single block.
Consensus on the combined set would depend on the tools that Alpenglow would provide; Votor’s fast off-chain voting and Rotor’s single-round data propagation - which is why MCP can’t come before Alpenglow. Anza’s design proposal for getting MCP onto Solana, dubbed Constellation, will have to figure out the mechanics: how to merge concurrent proposals, how to handle fees, and how to deal with conflicting transactions across concurrent proposals.
The key motivation for MCP is geographical. In the current design, every transaction in the world can flow to wherever the leader is, and traders colocate servers next to the validators the way high-frequency traders collocate in New Jersey in order to trade in front of one another. However, with leaders distributed across the globe, a trader can now see orders that go in New York at the same time that another order that goes in Tokyo, something that co-founder Anatoly Yakovenko has said is impossible for a single colocated server anywhere in the world to replicate.
How MCP Compares to Other Chains
Currently, every big chain runs single-proposer blocks. One Bitcoin miner wins every block, one Ethereum validator produces each 12-second block, and one Solana leader produces every 400-ms slot. Researchers have already started talking about the multiple-concurrent-proposer idea as a way to combat censorship on Ethereum, but nobody has built or deployed it. With Constellation, Solana would be the first large-scale chain to run concurrent block production, and that’s a double-edged sword: we don’t know how that system will go, and how we will be able to effectively combine concurrent proposals is a real problem that can lead to real engineering failures.
Censorship is cheaper to resist, too. In order to censor a transaction on Solana today, you only have to convince one leader per slot; with MCP, you would have to get all concurrent leaders per slot to agree. This is especially true in ecosystems like Solana that are increasingly hosting stablecoin payments and institutional flows via things like the Drift or Kamino protocols; the ability to censor is a security feature, not only a performance feature.
Why MCP Matters
MCP is Solana’s final goal, and part of what the roadmap calls the internet capital markets, where being included is a right not a privilege. For traders, MCP means better speed to the closest proposer and less of an advantage for those that collocated as much as they could. For users, MCP means you can still process your transaction, and it’s not up to one leader to keep you on chain. For an Ethereum user who’s skeptical about Solana’s ambitions and its ability to keep delivering on big-picture promises, it’s worth watching Constellation as it could be a sign that the 2025 to 2027 roadmap is realistic, Alpenglow is the start of the game not the finish of it.
