Solana's Alpenglow upgrade is now live on devnet, giving developers their first chance to test a consensus system built to cut transaction finality from about 12.8 seconds to roughly 150 milliseconds. Mainnet still runs on the current system, and Anza has set no firm activation date.
Solana's Alpenglow upgrade has reached the network's public developer environment, letting application teams test a system designed to cut transaction finality from about 12.8 seconds to roughly 150 milliseconds. Anza, which develops Solana's core validator software, announced the devnet switch on Sept. 25, one day after testnet completed its own transition.
Alpenglow changes how validators finalize blocks
Under the current system, TowerBFT, validators submit votes as transactions that appear inside blocks, and enough votes must accumulate across 32 slots before a block becomes final. That process currently takes about 12.8 seconds, according to the Solana Foundation's upgrade page.
Alpenglow's first phase, called Votor, has validators send votes directly to one another instead of through onchain transactions. The Foundation says a block can reach finality after one voting round if validators representing at least 80% of stake vote to accept it, with a second round providing another path when the first round falls short of that threshold.
Devnet lets application teams check their software against Alpenglow using tokens that carry no real value, while testnet gives validators and infrastructure operators a place to test the network under more demanding conditions. Solana's mainnet still runs TowerBFT, so the 150-millisecond figure remains a target rather than a finality time available to users today.
Block-data services must adapt before mainnet
Applications that only send transactions and read balances need no migration, the Foundation says, because transaction execution, fees, and transaction formats stay the same under the consensus upgrade. Services that build transaction histories face more work, however.
Alpenglow can expose competing candidate blocks for the same slot before the network settles on one, so the Foundation is telling data providers to keep those candidates separate and retain only the block that reaches confirmation. Validator votes will also disappear from blocks since they no longer travel as transactions, meaning a chart counting both user transactions and validator votes will show a lower total after activation even if user activity stays the same.
Mainnet activation still has no firm date
Solana's earlier roadmap tied a proposed mainnet rollout to Agave 4.3 and an October target, but neither the testnet transition nor the devnet activation sets a confirmed date for the live-network switch. Anza's software schedule tentatively allows mainnet feature activations to resume on Sept. 28, though it does not name that day as Alpenglow's activation date.
The Foundation describes Votor as Alpenglow's first phase. A later phase, Rotor, is planned to replace the system used to spread blocks across the network, while the roughly 150-millisecond target still comes from testing and simulations rather than transactions settled under live-market conditions.
Source: crypto.news
Trading involves risk.