This is a segment from the 0xResearch newsletter. To read full editions, subscribe.
Solana’s consensus stack, proof-of-history sequenced with TowerBFT, has always delivered faster block times than competitors. But finality still sits around 10-20 seconds, far from the Nasdaq-level latency Solana aspires to.
SIMD-0326 (“Alpenglow”) proposes cutting finality to ~100-150ms by moving validator voting off-chain. Today, validators continuously post votes onchain to signal fork choice, and these votes dominate throughput despite carrying no user value.
Source: SolScan
Under Alpenglow, validators instead pay a fixed “Admission Ticket” of 1.6 SOL per epoch, burned to the network. Leaders then gather votes off-chain through a component called Votor, compress them into certificates, and write those certificates onchain. The result is a swap: Millions of low-value vote transactions are replaced by one predictable fee per validator, lowering consensus overhead by ~20% and freeing blockspace for user activity.
The design also adjusts fault tolerance. TowerBFT today remains live unless more than 33% of stake is adversarial. Alpenglow introduces a “20+20” model, where the chain stays live with 20% malicious stake and another 20% offline. For applications like DeFi, specifically perpetual exchanges like Drift, sub-second finality transforms Solana from “fast” into a real-time settlement layer.
The economics are still under debate. Smaller validators face a flat 1.6 SOL per epoch fee regardless of stake, while reward flows remain undefined. Governance discussions (epochs 833-842) have emphasized the need for a clear rollout path, including sequencing for Alpenglow’s components. Still, if implemented, SIMD-0326 would represent one of Solana’s most significant structural upgrades.
Get the news in your inbox. Explore Blockworks newsletters:
Source: https://blockworks.co/news/solana-prepares-for-alpenglow