Anza's Alpenglow Upgrade Cuts Solana (SOL) Finality to 150 Milliseconds

Update 19:24 UTC: In an October 22 video post, Solana (SOL) Developers framed it as an adaptation under proposal, and the main Solana account said the network "has been early in quantum readiness."

Alpenglow Replaces Tower BFT on Mainnet

Solana's rebuilt consensus layer, Alpenglow, went live on mainnet-beta on Wednesday, October 21, replacing Tower BFT as the set of rules the network uses to lock in blocks. Anza, the developer behind the Agave validator client, announced the switch in a post on X, calling it the largest consensus change in the network's history. The upgrade is measured in milliseconds rather than in the Solana (SOL) price. A consensus mechanism is the process by which validators, the operators who secure the delegated proof of stake network, agree on transaction order and on finality, the point at which a payment counts as settled. Under Tower BFT, validators recorded votes as blockchain transactions and stacked approvals across 32 slots before a block became final, a wait the Solana (SOL) Foundation put at roughly 12.8 seconds. Alpenglow targets about 150 milliseconds. Its new voting protocol, Votor, lets validators exchange votes directly and bundle them into certificates, so a block can reach finality after one or two voting rounds. The handover itself was quick: Anza reports the switch from Tower BFT to Alpenglow took 900 milliseconds, and 446 of the 654 participating validators certified the genesis block that anchors the new consensus. Because the migration formed a fresh genesis block while keeping the existing chain, no user balances moved. Validators face a new cost model as well, a flat 0.8 SOL participation fee per epoch in place of per-vote transaction fees. An epoch is a bundle of slots, the network's operating cycle for scheduling changes like this one. Anza frames the benefit for institutions, from settlement with cryptographic guarantees to always-on capital markets, a case aimed most directly at firms in traditional finance, where settlement windows still run in days.

The upgrade appears to have lifted the market along with the network. In the hours after the switch, Solana (SOL) traded at $108.96, up 2.87% over 24 hours and the highest since October 14, an 8.6% recovery from the low of $100.32. Dashboards recorded finality at 6.6 seconds immediately after activation, with throughput at 4,721 transactions per second, while the migration was certified by validators holding 82% of staked stake and the network did not stop.

The gap between target and live performance also shows up against the testnet record, where Alpenglow's median finality was measured at 54 milliseconds, a figure that cannot be compared directly with mainnet because the validator environment differs. Market data placed Solana at $108.40 before the upgrade, a 3.11% gain over 24 hours, though the token remains down roughly 42% over the past year. Validators are also preparing a separate client change, Agave 4.5, which raises the number of accounts a single versioned transaction can support from 64 to 96; the Solana Foundation says this can improve asset exchange paths and execution pricing, and the upgrade is in development with a December 2026 activation target.

A day after activation, the upgrade has also registered outside English-language market coverage: a Japanese Web3 news wrap for October 22 listed Alpenglow's start of operation on mainnet-beta among the day's leading blockchain stories, describing the network as targeting finality of about 150 milliseconds. The item adds no new performance figure or schedule change, but its placement at the top of a daily roundup shows the consensus switch being tracked by regional media alongside unrelated protocol news such as Zcash's NU7 activation. For Solana watchers, it confirms the migration is being reported as a completed mainnet event rather than a testnet trial, with the 150-millisecond figure still framed as a target.

Solana's developer accounts used the day after the migration to pitch a follow-on research track: Quantumglow, a proposed post-quantum adaptation of Alpenglow that would protect validator agreement against quantum attacks while keeping fast finality. In an October 22 video post, Solana Developers framed it as an adaptation under proposal, and the main Solana account said the network "has been early in quantum readiness." Neither announcement set an activation date or identified a dedicated SIMD. The research itself is not new: Anza published its Quantumglow overview on July 30, naming Quentin Kniep, Kobi Sliwinski and Roger Wattenhofer as authors and pointing to Section 5 of the updated Alpenglow whitepaper. The design centers on Ax, short for Alpenglow XMSS, a hash-based signature scheme in which validators cache one-time public-key material ahead of use, paired with a signed block commitment for authentication.

@anza_xyz · X post

Post on X.

View on X

Users Keep Wallets, Dashboards Change

For everyday users, little changes on the surface. Wallets, application behavior, transaction formats and the fees ordinary users pay all stay as they were, so funds keep moving through existing apps and no action is required from anyone holding SOL. What improves is the point at which money is truly settled. Many applications already displayed transfers as “confirmed” before the network had finished locking them in, which made payments feel quick while a further wait for full settlement lingered. With Alpenglow, that display carries stronger backing: a block reaches finality in one or two voting rounds, so an exchange can credit a deposit sooner and let a customer trade with the funds, and a merchant accepting a stablecoin payment can confirm it without holding the customer at the counter. The change matters most where a service releases funds or goods the moment payment arrives, since the gap between a displayed and a settled balance was the main friction under the old design. Anza notes that some computers receive finality information later than the validators voting directly, so screen-level speed still depends on the service handling the payment, and the 150-millisecond figure is a target under normal network conditions rather than a guarantee for every transaction. Its framing extends to tokenized real-world assets, where faster finality shortens delivery-versus-payment windows. One visible side effect lands on analytics dashboards: validator votes no longer count as transactions, so services that tallied them will report lower totals, and explorers have been directed to reset throughput baselines at epoch 1064. Developers building across the Solana ecosystem gain the same settlement guarantee, whether they ship trading apps, payment rails or consumer products.

Phase One Done, Rotor Next

The two announcements describe one roadmap rather than two events. This mainnet switch is phase one of Alpenglow, and Rotor, a redesign of how block data propagates across the network, comes next, though Anza has set no date for it. Until that lands, the flat 0.8 SOL per-epoch fee is the number worth watching: it replaces variable per-vote costs with a fixed charge, and validator operators will see the effect at the next epoch boundary. Anza shipped the whitepaper fewer than 18 months before activation, a pace that suggests further steps may follow quickly once Rotor is scheduled.