# Alpenglow Consensus: What 100 ms Finality Means

<figure><img src="https://2187881347-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F9V0rc1Tyxf8VyxChx53w%2Fuploads%2F89eGjNWFF9dtIuyJbrhR%2FAlpenglow.png?alt=media&#x26;token=981c08b6-af66-4d34-b57c-5bd669b87959" alt=""><figcaption></figcaption></figure>

Solana’s next big leap is Alpenglow, a consensus overhaul designed by Anza and revealed at Solana Accelerate 2025. In simulations the new protocol trims block finality from today’s \~12.8 seconds to 100–150 milliseconds, about one-tenth of a blink. For anyone who stakes SOL or runs a validator, that speed-up carries real economic weight.

### How Alpenglow Works

| **Legacy Stack**                                                       | **Alpenglow Stack**                                                                                                                                                                 |
| ---------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Proof-of-History clock + Tower BFT (two-round voting) + Turbine gossip | <p>Votor – single-round fast vote (with a concurrent fallback round)</p><p>Rotor – one-hop shred relay, 400 ms slot clock</p><p>Votes bundled into lightweight BLS certificates</p> |

* Votor finalises a block the instant ≥ 80 % of stake votes; if that threshold isn’t reached, a concurrent second round completes at ≥ 60 %.
* Rotor pushes shreds through dedicated relay nodes rather than multi-hop gossip, so data and votes arrive almost simultaneously.
* All votes move off-chain and are bundled into certificates, eliminating per-slot vote transactions and their fees.

### Why Sub-Second Finality Matters for Stakers

| **Impact Area**     | **What Changes**                                                          | **Benefit to Stakers**                                                  |
| ------------------- | ------------------------------------------------------------------------- | ----------------------------------------------------------------------- |
| Reward Stability    | Fewer orphaned slots and quicker replay                                   | Smoother, more predictable staking APY                                  |
| Fee Flow            | Apps can settle trades in real time → higher TPS and priority-tip bidding | More fee revenue shared by Jito and other MEV clients                   |
| Validator Economics | Voting costs (today \~25 % of gross rewards for small nodes) disappear    | Smaller operators keep a larger share, encouraging decentralisation     |
| User Adoption       | Payments, games and social feeds feel Web2-fast                           | Sustained demand supports SOL price, indirectly lifting staking returns |

### What Validators Need to Know

* Performance headroom – Benchmarks show \~25 % lower average CPU idle time because buffers recycle sooner.
* No exotic hardware – Alpenglow keeps the 400 ms slot; existing rigs simply upgrade binaries after the SIMD vote.
* Roll-out plan
  * Devnet & security audits: H1 2025
  * Public testnet (fault-injection, relay stress): Q3 2025
  * Mainnet feature-gate vote: late 2025
  * Target general activation: early 2026 once two-thirds of stake opts in

#### Toby’s View

Near-instant finality doesn’t just make Solana feel snappier, it rewires validator economics. Lower operating costs and richer priority-tip markets align perfectly with Toby’s mission: optimising validators and unlocking Open MEV so yield flows efficiently to stakers. We’re already benchmarking Alpenglow testnets to ensure our optimiser is day-one ready.

<br>