Solana’s Alpenglow Upgrade Records 54ms Median Finalization on Testnet
Solana’s Alpenglow consensus overhaul has reached a 54-millisecond median finalization time on public testnet, beating its design target and setting the stage for a potential mainnet deployment.
Solana’s Alpenglow upgrade is now live on the network’s public testnet, where it has finalized blocks in a median of 54 milliseconds. The result substantially outperformed the upgrade’s 100–150 millisecond design target and could mark a major improvement over Solana mainnet’s previous approximate finality time of 12.8 seconds.
Testnet Deployment and Performance
Alpenglow activated on Solana’s public testnet on September 24, 2026, at slot 444,625,255, following a 17-minute countdown. The upgrade reached devnet the next day at slot 504,148,999, giving developers additional environments in which to evaluate the consensus overhaul.
The headline change is Votor, a new consensus mechanism that replaces TowerBFT, Solana’s voting system since the network launched in 2020. The 54ms median recorded on testnet is faster than the low end of Alpenglow’s original performance goal.
- Testnet median finalization: 54ms
- Alpenglow design target: 100–150ms
- Previous approximate mainnet finality: 12.8 seconds
How Alpenglow Changes Consensus
Under Solana’s previous model, validators submitted votes as on-chain transactions. Those votes typically consumed between 50% and 75% of block space, creating a significant use of network capacity.
Votor moves the voting process off-chain. Validators exchange BLS-signed votes directly through a peer-to-peer network, then aggregate those votes into certificates. Because BLS signatures can bundle multiple validator approvals into a compact proof, the approach is designed to reduce overhead while accelerating agreement.
Alpenglow also introduces dual finalization paths, giving blocks two potential routes to finality, along with a “20+20” fault-tolerance model intended to improve resilience and performance. The upgrade does not alter the Solana Virtual Machine, transaction formats, or fee structures, which should limit the changes required from application developers and users.
The proposal was formalized as SIMD-0326 and approved in a September 2025 validator vote with 98.27% support. Stake participation in that vote was 52%.
What This Means
The immediate test for Alpenglow is whether its testnet performance can translate to mainnet conditions. Mainnet activation has no fixed date and is tied to the Agave 4.3 release, expected in late September or October 2026, subject to developer observations from testnet and devnet.
A mainnet result near 54ms would exceed expectations, but even performance within the 100–150ms target range would represent a dramatic improvement over approximately 12.8 seconds. Developers will watch the Agave release schedule, testnet stability, and any issues that could delay deployment before assessing Alpenglow’s impact on Solana’s competitiveness for latency-sensitive applications.