Solana switches to Alpenglow, targeting 150ms finality

Solana activated Alpenglow on mainnet on October 21, replacing the TowerBFT consensus system it had used since launch with a new voting protocol that targets transaction finality in about 150 milliseconds, down from roughly 12.8 seconds. The Solana Foundation says the change leaves transaction execution, fees, wallets and apps as they were. In its Alpenglow upgrade notes, the Foundation says the protocol went live at the start of epoch 1064, at approximately 13:36 UTC.

That target concerns finality: the point at which an application can treat a transaction as settled and no longer reversible. It is distinct from the earlier “confirmed” status that many apps displayed while the network still needed time to lock a transaction in. CoinDesk reported on the upgrade’s user effects, including the possibility of faster exchange deposits and stablecoin payments. The practical gain depends on services acting on finality information quickly; 150 milliseconds is a target under normal conditions, not a guaranteed wait for every user.

How does Alpenglow change the wait for settlement?

Under TowerBFT, validators recorded their votes as transactions and accumulated approvals across successive blocks. Alpenglow’s Votor protocol instead lets validators send votes directly to one another and combine them into certificates. The Foundation says blocks can then become final after one or two rounds of voting, rather than waiting for votes to build up over as many as 32 slots under the old system.

For a user, this could make a payment feel closer to an instant card transaction when a merchant or app waits for finality before releasing goods or funds. An exchange could credit a deposit sooner, while a shop could receive stronger settlement assurance before the customer leaves. That assurance is different from a fast on-screen confirmation: the app has to use the finality signal, and its connection to the validators affects how quickly it receives that signal.

What stays the same, and what does the change cost?

Sending a transaction does not require a new wallet or a different fee. Solana says the upgrade does not change its SVM execution environment, transaction formats, programs or fee mechanics. Users who only send transactions and check account balances do not need to migrate. That keeps the change narrower for everyday users than a redesign of how applications execute transactions would be.

The trade-off is more substantial for validators and infrastructure. Mainnet validators needed an Agave-based client, version 4.3.2, for the migration; the Foundation says Firedancer and Frankendancer were not supported at that point. Voting validators also need at least 0.8 SOL above the rent-exempt minimum in their vote account to pay the Validator Admission Ticket fee each epoch. Developers and data services that read block streams or count transactions may need to adjust: votes no longer appear as blockchain transactions, and some streaming data has changed.

That removal also changes how activity looks in dashboards. A reported transaction total may fall because validator votes are no longer counted, even if users continue making the same number of payments and trades. Historical comparisons will need to distinguish user transactions from votes that previously occupied blocks.

Which signals will show whether users benefit?

Watch real-world finality times across ordinary network conditions, and whether exchanges, merchants and apps begin acting on the faster settlement signal. The 150-millisecond figure is a target, so the gap between validator finality and what users see in an app will matter. Infrastructure providers’ handling of the new block and vote data is another test: dashboards that fail to separate validator votes from user activity could make network usage appear to change when only the measurement has changed.

Alpenglow’s rollout also has a later phase. The Foundation says Rotor, intended to replace Turbine’s block-data distribution protocol, is planned for a future release but is not yet scheduled. For now, the clearest measure of Alpenglow’s effect is whether faster finality reaches applications reliably without creating confusion in the activity data used to judge the network.

Sources and documents

• Alpenglow upgrade notes — solana.com
• reported on the upgrade’s user effects — coindesk.com