What Is Alpenglow in Solana

TL;DR

• Votor replaces Tower BFT voting and removes vote transactions from blocks.
• Rotor will replace Turbine in a later, unscheduled phase.
• The 150 ms figure is a target, and mainnet performance is still unmeasured.
• Transaction execution and fees stay unchanged.
• Alpenglow activates on Solana mainnet-beta in epoch 1064, which starts on Wednesday, October 21, 2026, at approximately 15:00 UTC.

Alpenglow is Solana‘s new consensus protocol and the first full replacement of Tower BFT. It targets transaction finality of about 150 ms, down from 12.8 seconds, according to the Solana Foundation.

Anza unveiled it in May 2025 in a technical whitepaper.

Alpenglow activates on Solana mainnet-beta in epoch 1064, which starts on Wednesday, October 21, 2026, at approximately 15:00 UTC.

Anza expects the switch from Tower BFT to Votor around 14:50 UTC and says it will complete within a single slot, after 18 months of development.

Epoch start times drift with slot timing, so the epoch number is a more reliable reference than the clock. Activation also requires 95% of stake to run the new software for two consecutive epochs.

What Is Alpenglow and Why Does Solana Need It?

Alpenglow is designed to cut finality from 12.8 seconds to about 150 ms and remove vote transactions from blocks. Its core proposal, SIMD-0326, passed governance on September 2, 2025, with 98.27% of voting stake in the final tally.

How Solana Reaches Consensus Today

Under Tower BFT, validators vote through on-chain transactions, and a block is final after votes accumulate over 32 slots. Proof of History (PoH), a continuous SHA-256 hash chain, paces block production within each slot.

Consensus moves each block through three stages:

• Block production: a leader builds a block in its 200 ms slot.
• Confirmation: a supermajority of stake votes for the block, usually within one slot.
• Finality: the block becomes irreversible, after about 12.8 seconds under Tower BFT.

What Alpenglow Is Designed to Improve

Alpenglow targets three limits of Tower BFT:

• finality latency of 12.8 seconds,
• ledger growth from vote transactions,
• operational complexity for validators.

The aim is real-time responsiveness for DeFi and payments.

How Does Alpenglow Work?

Alpenglow has two components: Votor, a consensus protocol built on direct validator voting, and Rotor, a block propagation layer. Transaction execution is untouched.

Votor: Solana’s New Voting and Consensus Protocol

Votor replaces on-chain voting with votes that validators send directly to each other, aggregated into BLS-signed certificates. Vote transactions and their fees leave the block.

Each proposed block follows one of three routes:

• Fast path: at least 80% of stake notarizes in round one, and the block is final.
• Slow path: 60% to 80% notarizes, and a second round at 60% finalizes the block.
• Skip: 60% of stake votes to skip the slot.

Network latency and topology decide which path wins. According to the Solana Foundation, a tightly clustered 60% of stake can finish two rounds before a spread-out 80% finishes one.

Rotor: A New Block Propagation Mechanism

Turbine distributes block data through a multi-layer tree of validators. Rotor will replace it with a single relay layer while keeping erasure coding and stake-adjusted bandwidth.

Rotor follows the Votor rollout, Anza expects the switch from Tower BFT to Votor around 14:50 UTC and says it will complete within a single slot, after 18 months of development.

What Happens to Proof of History and Tower BFT?

Votor replaces Tower BFT‘s voting and lockouts, and its timeouts and certificates take over block pacing from Proof of History. Leaders stop emitting tick entries, and the block leader sets the Clock timestamp within a bounded range.

Transaction ordering and execution continue as before.

A block is final after one round when 80% of stake notarizes it and after two rounds at 60%; a slot is skipped when 60% votes to skip. Rotor changes only how block data reaches validators and ships in a later release.

What Are the Main Benefits of Alpenglow?

Alpenglow brings three benefits:

• faster finality,
• lower on-chain voting overhead,
• a different fault-tolerance model.

Faster Transaction Finality

Votor targets about 150 ms finality, a figure that depends on stake distribution, against 12.8 seconds under Tower BFT.

Block time is a separate figure: 200 ms after the reduced slot times upgrade. Under Alpenglow, the confirmed and finalized commitment levels merge into one state.

It’s been reported to reach a 54 ms median on testnet. Mainnet results will depend on its larger, more spread-out validator set.

Lower On-Chain Voting Overhead

Votes travel directly between validators, so blocks hold only user transactions. TPS dashboards will drop sharply at activation while real user activity stays the same.

Changes to Fault Tolerance

Votor stays live with 20% adversarial plus 20% offline stake, versus about 33% faulty stake under Tower BFT. The Anza whitepaper proves its safety under these assumptions, and outages and software bugs remain possible.

Tower BFT vs. Alpenglow at a glance

Metric	Tower BFT (mainnet today)	Alpenglow (Votor)	Figure type
Transaction finality	~12.8 s	~150 ms	Current measurement vs. design target
Finality on testnet	Not applicable	54 ms median (reported)	Testnet observation
Block (slot) time	200 ms	200 ms (unchanged by Alpenglow)	Current measurement
Voting method	On-chain vote transactions	Direct votes aggregated into BLS certificates	Protocol design
Validator voting cost	Vote fees, about 2 SOL per epoch	Flat 1.6 SOL VAT per epoch, burned	Protocol design
Commitment levels	processed / confirmed / finalized	confirmed = finalized	Protocol design
Fault tolerance	~33% faulty stake	20% adversarial + 20% offline	Protocol design
Block propagation	Turbine	Turbine now, Rotor later	Planned

How Will Alpenglow Affect Solana Validators and Developers?

Most users and applications need no action. The changes fall on validator operators and on infrastructure that reads block data.

What Changes for Validators?

Validators exchange BLS-signed votes directly and prepare in three steps:

1. Register a BLS public key on the vote account.
2. Upgrade to Agave 4.3 on Anza‘s release schedule.
3. Run only one active node to avoid equivocation from a primary and hot spare.

The Validator Admission Ticket (VAT) has been active on mainnet since July 22, 2026, excluding validators without a BLS key and capping the set at 2,000. Once Alpenglow is live, each validator pays a flat, burned 1.6 SOL per epoch, close to today’s voting cost of about 2 SOL.

Delegators staking SOL with Everstake or any other validator need no action.

What Changes for Developers and Infrastructure Providers?

Applications that only send transactions and read account state have nothing to migrate. Infrastructure that reads blocks needs these updates:

• Geyser and gRPC consumers: buffer events per (slot, bank_id) and keep only the bank that reaches Confirmed.
• Multi-provider pipelines: reconcile streams on the blockhash, because bank_id is node-local.
• Indexers and dashboards: re-baseline transaction counts and TPS, since vote transactions leave blocks. Read validator participation from block footer certificates instead of Vote program instructions.
• Time-sensitive logic: tick entries disappear, and the leader sets Clock.unix_timestamp within bounds.
• RPC clients: after activation, move new code to finalized, because confirmed will be removed later.

The getAgGenesisCert RPC method responds with null until a cluster migrates.

What Does Alpenglow Mean for Applications and Institutions?

Finality near 150 ms could support point-of-sale payments and trading. Institutions that need irreversible settlement, including users of Everstake institutional staking, may see the largest change once mainnet performance is proven.

When Will Alpenglow Launch on Solana Mainnet?

Alpenglow is expected to go live on Solana mainnet-beta in epoch 1064, which starts on Wednesday, October 21, 2026, at approximately 15:00 UTC.

Alpenglow’s Current Deployment Status

Agave 4.3 activated Alpenglow on testnet on September 24, 2026, and its validator prerequisites are already live on mainnet.

Date	Milestone	Cluster
Oct 2026	Mainnet activation expected in Q4 2026; the Agave v4.4 schedule resumes mainnet feature activations on November 9, reportedly without Alpenglow	Mainnet
Sep 28, 2026	Mainnet feature activations resume under the v4.3 schedule; Alpenglow not activated	Mainnet
Sep 25, 2026	Alpenglow activated on devnet	Devnet
Sep 24, 2026	Alpenglow activated on public testnet	Testnet
Sep 18, 2026	Agave v4.3.0 released, including Votor	All clusters
Aug 5–19, 2026	Up to 50,000 SOL bug bounty competition	Off-chain
Jul 22, 2026	Validator Admission Ticket (SIMD-0357) activated	Mainnet
Jul 8, 2026	BLS public key management (SIMD-0387) activated	Mainnet
Sep 2, 2025	SIMD-0326 passes governance with 98.27% yes and 52% of stake voting	Mainnet governance
May 2025	Alpenglow announced with Anza whitepaper	Off-chain

The getAgGenesisCert RPC method is the most reliable activation signal.

Alpenglow follows other 2026 protocol changes such as SGP-0002 (Annual disinflation rate), approved on August 28, 2026.

When Will Rotor Be Implemented?

Rotor ships after Votor adoption, it is expected to go live on October 21, 2026, at approximately 14:50 UTC.

What Are Alpenglow’s Risks and Limitations?

Alpenglow‘s main risks come from the scale of replacing consensus on a live network.

Security and Fault Tolerance

Votor‘s properties hold only under its stated assumptions of 20% adversarial and 20% offline stake. Off-chain voting also removes fee-based throttling, so spam resilience is untested under mainnet load.

A malicious leader can push the Clock timestamp forward within a bounded range.

Implementation and Migration

Bugs or edge cases in Votor could surface only on mainnet. Anza ran an up to 50,000 SOL bug bounty and a public testnet activation before mainnet.

Downstream, a Geyser pipeline keyed on slot alone can merge competing blocks, and dashboards that count votes will misreport activity.

Decentralization and Network Performance

The Validator Admission Ticket caps the set at 2,000 validators, and a flat 1.6 SOL fee weighs more on low-stake operators. Finality also depends on how fast 80% of stake responds, so geography and topology will influence real-world results.

FAQ

What is Alpenglow in Solana?

Alpenglow is Solana‘s new consensus protocol that targets 150 ms finality using direct validator votes.

How fast will Solana transactions finalize with Alpenglow?

Alpenglow targets about 150 ms finality, down from 12.8 seconds.

What is the difference between Votor and Rotor?

Votor handles voting and finality, while Rotor handles block propagation. Votor ships in Agave 4.3, and Rotor replaces Turbine later.

Does Alpenglow replace Proof of History?

Yes, as the block pacing mechanism. Leaders stop emitting the 64 tick entries per slot that PoH produces, while ordering and execution continue on validators.

Does Alpenglow change Solana transaction fees?

No, user transaction fees stay the same. Validators pay a flat 1.6 SOL per epoch that replaces vote fees, and Everstake delegators need no action.

When will Alpenglow be activated on mainnet?

Alpenglow is expected to go live on epoch 1064.

Do Solana validators need to upgrade for Alpenglow?

Yes. Validators need a registered BLS key, required since July 22, 2026, and Agave 4.3 or later.

Does Alpenglow require changes to Solana applications?

Most applications need no changes. Apps that only send transactions, including Everstake staking integrations, keep working after Agave 4.3 activates Alpenglow, while indexers must handle bank_id.