Solana Eyes 100x Speed Leap — If It Dares To Kill Proof Of History

Solana, the blockchain that built its brand on a cryptographic stopwatch called Proof of History, is preparing to discard that mechanism in favour of a radically simpler design dubbed “Alpenglow.” Revealed in a 35-minute technical post by infrastructure firm Helius on May 26, the proposal claims nothing less than a two-order-of-magnitude leap in settlement speed: “Alpenglow’s headline benefit for Solana is a 100× reduction in time to transaction finality… allowing Solana to become competitive with more centralized Web2 infrastructure and enabling real-time applications.”

Solana’s Fastest, Boldest Consensus Rewrite Yet

Alpenglow is the work of Anza’s new research division under ETH Zurich’s Professor Roger Wattenhofer, whose 2024 paper exposed liveness flaws in the incumbent protocol. Asked to keep Turbine’s data-broadcast spine while proving the result correct, Wattenhofer’s team pursued a ruthless minimisation strategy. “We want to have the simplest possible protocol. Performance is number one for us when we develop a protocol, but simplicity is also important,” Wattenhofer explains in the blog post.

Simplicity here means collapsing three moving parts—Proof of History, Tower BFT and gossip-based vote propagation—into two. Rotor abandons Turbine’s multi-hop fan-out for a single-hop relay in which stake-weighted nodes push erasure-coded shreds directly to peers. Votor, the new consensus core, eliminates on-chain vote transactions entirely: validators emit single-packet votes that any node may aggregate into a BLS certificate once the weight crosses a threshold. A first-round quorum of 80 percent finalises instantly; failing that, a second 60 percent quorum completes finality inside the same 400 millisecond slot window.

Because votes never touch the ledger, roughly three-quarters of Solana’s present transaction flow disappears. “By eliminating these fees through off-chain voting, participation costs are dramatically reduced, making it more feasible for smaller validators to operate and lowering the barrier to entry for contributing to network security,” the authors note. Independent estimates suggest the stake needed for break-even operation falls from about 4,850 SOL to 450 SOL.

Alpenglow’s security budget shifts from the classical 33 percent Byzantine threshold to what the designers call a “20 + 20” envelope: “It maintains safety if up to 20 % of stake is controlled by adversaries, and liveness if an additional, separate 20 % of stake is offline or unresponsive.” The overlap between quorums ensures that no two conflicting blocks can both finalise without provably slashable double-signing by at least one-fifth of stake.

For node operators, vote-fee relief is only the beginning. Off-chain certificates shrink ledger growth and dissolve the traditional distinction between “confirmed” and “finalised” commitment levels; client logic that waits for two confirmations can now watch a single certificate. Yet the upgrade also shifts bandwidth patterns. Poll-based APIs become obsolete: a block that settles in 100–150 ms demands real-time WebSocket fan-out, and caches nursing state longer than a quarter-second risk going stale.

Because Rotor and Votor are agnostic to leader count, they open the door to long-mooted multi-leader execution. Co-founder Anatoly Yakovenko has floated an initial prototype in which two Alpenglow instances share the same relay set, emitting parallel shred streams. Execution-layer puzzles—conflict resolution, fee-market design, cross-lane MEV—remain unsolved, but the consensus blockers are gone.

The white-paper omits the mechanics of slashing and of compensating relays for bandwidth, leaving validator economics partly undefined. Still, governance headwinds seem modest: the proposal cuts costs for smaller validators rather than raising them. Helius pegs main-net deployment at “early next year,” contingent on a Solana Improvement Document, community review and an on-chain vote.

A latency histogram in Anza’s simulations captures the ambition: “Alpenglow’s finality overall is 2× the lower bound,” the authors write. In other words, the entire consensus overhead is twice the raw internet delay between a leader and two-thirds of stake. Achieving that in production would crown Solana the first global blockchain to deliver deterministic, sub-second finality without a central sequencer—proof, perhaps, that a network once caricatured as “speed at all costs” can grow up without slowing down.

At press time, SOL traded at $173.88.