PrefaceOn October 21, 2026, Solana changed the system that decides when a transaction is final. The Solana Alpenglow upgrade replaced TowerBFT, the voting method used since 2020, with a new design buiPrefaceOn October 21, 2026, Solana changed the system that decides when a transaction is final. The Solana Alpenglow upgrade replaced TowerBFT, the voting method used since 2020, with a new design bui

Solana’s Alpenglow Upgrade Goes Live, Targeting 150ms Transaction Finality

#Industry Perspectives

4$0.01253+10.68%

NOT$0.0003868+4.45%

GAIN$0.0014015+4.13%

Author: Emmanuel Olamiye

2026/10/28 16:00

Preface

On October 21, 2026, Solana changed the system that decides when a transaction is final. The Solana Alpenglow upgrade replaced TowerBFT, the voting method used since 2020, with a new design built by Anza. Its goal is 150ms finality, meaning a transaction becomes impossible to reverse in about a seventh of a second. TheSolana Foundation says it was activated in epoch 1064, at about 13:36 UTC.

Much of the coverage skips the details that matter. The 150ms figure is a target. The old 12.8-second baseline is outdated. And the second phase of Alpenglow is not live yet. This guide explains the upgrade step by step.

Key Takeaways

Alpenglow went live on Solana mainnet on October 21, 2026, replacing TowerBFT.
Its first phase, Votor, lets validators vote directly, targeting about 150ms finality.
Finality means a transaction cannot be reversed, which is different from block speed.
Early readings ranged from about 6.6 seconds at launch to 202ms later, with no sustained average yet.
Measured against today's faster block timing, the gain is closer to 40x than the advertised 85x.
Many apps already used a fast but weaker confirmation, so the biggest change is a stronger guarantee at similar speed.
Users need to do nothing, while validators had to switch software and pay a flat fee.
Rotor, the second phase, has no release date, and SOL did not break out after launch.

Key Terms

Validator: a computer run by an independent operator that checks transactions and helps secure Solana.
Stake: SOL locked up behind a validator, which gives it voting weight.
Slot: a short window, now about 0.2 seconds, in which one validator builds a block.
Epoch: a longer period, roughly a day at current speeds, used to organize validator duties and fees.
Finality: the point where a transaction can no longer be undone.
Certificate: a combined, signed proof that enough validators agreed.
TowerBFT, Votor, Rotor: the old voting system, the new voting system, and the planned new block-delivery system.

1. What Is the Solana Alpenglow Upgrade?

Alpenglow is Solana's new consensus protocol; the rules validators follow to agree on transactions. Its first phase, Votor, lets validators send votes straight to each other instead of recording them as transactions, so a block can be final in one or two voting rounds. It targets about 150 milliseconds, down from about 12.8 seconds. Wallets, apps, and fees do not change.

1.1. Why TowerBFT Was Slow

Under TowerBFT, validators voted by sending vote transactions. Each vote had to be written into a later block. A block counted as final only after votes had stacked up over 32 slots, according to the Solana Foundation. At the old 400ms slot length, that took about 12.8 seconds.

Those votes also filled much of every block and cost validator fee, which made the system heavy as well as slow.

1.2. What Alpenglow Changes in Solana Consensus

Alpenglow replaces the voting layer of Solana consensus. Validators send votes directly to each other and combine them into certificates.

What changes: how blocks are agreed on, how fast they become final, how votes are sent, and what validators pay.

What stays the same, according to the Foundation: the Solana Virtual Machine, programs, transaction formats, fees, wallets, and the account model. Slot timing was changed separately.

1.3. Who Built It and How It Got Here

Anza, a Solana engineering team spun out of Solana Labs, built Alpenglow, according to1inch. Its research team, led by ETH Zurich professor Roger Wattenhofer, published the design in May 2025, perBigGo Finance. Anza CEO Brennan Watt called it the most ambitious blockchain upgrade since Ethereum's Merge, asCrypto.com reported.

May 2025: Anza researchers publish the Alpenglow design.
September 2, 2025: Validators approve it with 98.27% of votes, with about 52% of stake taking part, percrypto.news.
May 2026: It reaches Anza's community test network.
September 25, 2026: It activates on devnet.
October 21, 2026: It activates on mainnet.

New to the network? Start with thiscomplete beginner's guide to Solana.

2. How Does Alpenglow Work?

2.1. A Transaction Before and After Alpenglow

Here is the journey of a transaction under TowerBFT:

You send a transaction, and the current leader validator includes it in a block.
Other validators check the block and send vote transactions.
Those votes land in later blocks and pile up.
After about 32 slots of votes, your block is final.

And here is the same journey under Alpenglow:

You send a transaction, and the leader includes it in a block, exactly as before.
Validators check the block and send their votes directly to each other.
If validators holding 80% of stake approve, the block is final after one round. If support is between 60% and 80%, a second round finalizes it.
A certificate proves the result, targeting about 150 milliseconds in total.

Whichever route reaches its threshold first wins. Which one that is depends on network speed and where stake is located, not on your transaction.

2.2. What Happens When Something Goes Wrong

Validators can also vote to skip a block that never arrived, and 60% of stake voting skip is enough to move on. As a further safety valve, validators progressively extend block times if the network cannot finalize a block for 10 seconds, giving messages more time to arrive, according to areport on Anza's team.

2.3. The 20+20 Safety Rule

Alpenglow keeps working if up to 20% of stake is malicious and another 20% is offline. Older designs tolerated up to one-third malicious stake, so this is a trade-off. Supporters argue that counting offline validators better matches real-world failures, per1inch.

2.4. Votes Leave the Blockchain

Because votes are no longer transactions, they vanish from blocks. Solana Compass data put them at about 62% of all transactions in the week before the switch.

Validator costs change too. The Foundation says voting cost about 1 SOL per epoch in vote fees, andBigGo says that roughly doubled after slot times were halved. Alpenglow replaces vote fees with a flat 0.8 SOL per epoch, which is burned.

3. How the Mainnet Switch Happened

Changing a blockchain's core rules while it keeps running is rare. The upgrade switch turned on at the start of epoch 1064, but the network kept using TowerBFT for another 5,000 slots, about 18 minutes. The last block in that window, slot 459,652,999, became Alpenglow's starting point.

Validators then signed a certificate for it. The migration needed validators holding 82% of stake to take part, perFinanceFeeds, and 446 of 654 validators signed. Alpenglow took over at slot 459,653,000, which Anza announced at 13:54 UTC. The Foundation says the handoff took about 900 milliseconds, andCrypto.com reported no outages.

Readiness was a worry beforehand. On October 19, validator operator Pumpkin's Pool flagged 45.2% of stake as not fully ready, though a large part ran Agave versions Anza said could still migrate. An October 20 snapshot flagged 20.3%. FinanceFeeds notes the two figures are not directly comparable.

4. Does Solana Really Reach 150ms Finality?

The Foundation calls 150ms an expectation, not a constant. Early readings have varied:

The Foundation's dashboard showed about 6.6 seconds shortly after activation, perTradingView News.
It showed 202ms on October 21 and 22, according toAltcoin Buzz, which noted this was one reading, not an average.
Anza lead economist Max Resnick has said users should see confirmation latency of about 150 to 200 milliseconds, perKuCoin.

No sustained, independent average exists yet, so treat 150ms as promising but unproven. There is another caveat. Anza's operator documentation says nodes outside the admitted validator set may learn about finality later, by up to one block interval, percrypto.news. So the 150ms describes when validators reach finality, and some users may see it slightly later.

4.1. The Practical Gain: A Stronger Guarantee

Here is the point most coverage misses about Solana finality. Under TowerBFT, the Foundation says almost everyone read the confirmed status, which arrived in about one slot. It was fast but weaker, since a confirmed block could in principle still be dropped. Certainty meant waiting for finality.

Under Alpenglow, confirmed and finalized now mean the same thing. Apps can take the stronger guarantee at close to the same speed. For many apps, the real gain is certainty, not raw speed.

4.2. Finality Is Not Block Time or TPS

Finality is when a transaction cannot be reversed. Block time is how often blocks appear, and TPS counts transactions per second. Alpenglow changes only the first. For comparison, Ethereum needs about 13 minutes for full finality, per BigGo. To follow transactions yourself, see thisSolana explorer guide.

5. What Changes for Users, Developers, and Validators

Everyday users: Nothing. Signing, sending, and fees work as before. If you use MetaMask on Solana, see our guide onwhether MetaMask supports Solana.

Developers: Apps that only send transactions need no changes. Teams that track blocks should use the finalized setting and reset transaction-count dashboards, since votes no longer appear in blocks. The lower TPS is an accounting change. Solana Compass saw other activity hold steady at about 2,150 to 2,500 transactions per second around the switch.

Solana validators: Operators had to run version 4.3.2 of Agave, Jito-Solana, or Harmonic Salsa, register a new key, and keep 0.8 SOL above the minimum in their vote account. Missing the fee means no voting or rewards the next epoch. Firedancer and Frankendancer were not supported during the switch. Crypto.com reports Firedancer's latest release now supports Alpenglow, but Frankendancer will not.

6. Why Faster Finality Matters

Finality is when an exchange, bank, or merchant can treat a payment as complete. A shorter wait could let shops release goods sooner and exchange credit deposits faster. The Foundation points to near-instant confirmation at checkout as one example.

That makes Alpenglow relevant to payments, where Solana already carries stablecoin activity. Ourstablecoin beginner's guide explains why. Still, gains depend on whether apps, exchanges, and merchants change their own rules, and none have announced changes yet.

Conclusion

The Solana Alpenglow upgrade is the biggest change to the network's core since launch, and the switch itself went smoothly. A new consensus engine took over a live network in under a second, and wallets kept working.

The key claim still needs evidence. The 150ms figure is a target, early readings were mixed, the real speed gain is closer to 40x than 85x, and the clearest benefit may be a stronger guarantee at similar speed. Rotor is still to come. Watch sustained finality data over the next few weeks. Meanwhile, review theSolana price history before making decisions.

Frequently Asked Questions (FAQ)

1. What is the Solana Alpenglow upgrade?

The Solana Alpenglow upgrade is a rewrite of the rules validators use to agree on transactions. It replaced TowerBFT on October 21, 2026, and targets about 150 milliseconds to finality, down from about 12.8 seconds. It changes how validators vote and confirm blocks, but it does not change wallets, apps, or fees.

2. Is Alpenglow live on Solana mainnet?

Partly. The Foundation says Alpenglow activated in epoch 1064, and the new consensus took over at slot 459,653,000. That covers Votor, the voting phase. Rotor, the second phase that changes how block data is delivered, is not live and has no release date, so the full design is still unfinished.

3. How does Alpenglow work?

Validators send votes straight to each other and combine them into signed certificates. If validators holding 80% of stake approve a block, it is final after one round. If support is between 60% and 80%, a second round finalizes it. This removes the old step of recording votes as transactions, which frees up block space.

4. Does Alpenglow make Solana faster?

It makes certainty faster, but not everything else. Finality is the wait until a transaction cannot be reversed, and Alpenglow shortens it. Block time, about 200ms after a separate October change, is unaffected, as are fees and program execution. You wait less for certainty, not less for inclusion in a block.

5. Is 150ms finality guaranteed?

No. The Foundation calls 150ms an expectation that depends on network conditions and how stake is spread across validators. Early dashboard readings ranged from about 6.6 seconds right after launch to 202ms later, and no sustained average exists. Wait for several weeks of data before treating 150ms as proven.

6. What is the difference between Alpenglow and TowerBFT?

TowerBFT made validators send votes as on-chain transactions, and a block was final after about 32 slots of votes, or 12.8 seconds at the old slot length. Alpenglow sends votes directly between validators and finalizes in one or two rounds. It also replaces per-vote fees with a flat 0.8 SOL fee per epoch.

Disclaimer: This article is for information only and is not financial, legal, or tax advice. Crypto assets are volatile and carry risk. Network data and prices change quickly, so verify current figures before acting. Do your own research and consult a qualified professional before making decisions.

Market Opportunity

4 Price(4)

$0.01253

$0.01253$0.01253

USD

4 (4) Live Price Chart

The articles shared on this page are sourced from public platforms and are provided for reference only. They do not represent the position or views of MEXC. All rights belong to Emmanuel Olamiye. If you believe any content infringes upon the rights of a third party, please contact [email protected] for prompt removal. MEXC does not guarantee the accuracy, completeness, or timeliness of any content and is not responsible for any actions taken based on the information provided. The content does not constitute financial, legal, or other professional advice, nor should it be interpreted as a recommendation or endorsement by MEXC. For expert insights and in-depth analysis, visit MEXC Learn.

Latest Updates on 4