Solana’s Alpenglow Upgrade Launches on Testnet, Aiming for 150ms Transaction Finality

Solana’s Alpenglow upgrade is now live on its public developer network. That’s a meaningful step — it means real application teams can finally hammer on the code before it ever touches the main chain where actual money moves.

The core promise is pretty dramatic. Alpenglow aims to cut transaction finality from 12.8 seconds down to roughly 150 milliseconds. That’s not a modest tweak — it’s basically an 85x speed improvement, at least on paper. Anza, the company that builds and maintains Solana’s core software, made the announcement after the upgrade cleared a separate test network first. The Solana Foundation has since confirmed Alpenglow is running on both the devnet and the testnet simultaneously.

Fast. Really fast.

The devnet is where developers test applications using tokens that carry no real value — it’s a sandbox, essentially. The testnet is a harder environment, used mostly for stress-testing network software and validator operations under load. Getting through both of those before touching mainnet is the standard path, and Alpenglow is now on that path.

How Validators Actually Change

The technical shift here is more fundamental than most upgrades. Right now, validators — the machines responsible for verifying and agreeing on transaction records — cast their votes by recording them as transactions inside blocks. That takes up block space and adds steps to the process. Under Alpenglow, validators exchange votes directly with each other, cutting out the need to log those votes as on-chain transactions entirely.

The result, according to the Foundation, is that the network should only need one or two voting rounds to reach consensus. That’s a significant compression of the current process. And it’s where most of the speed gain comes from.

But there’s a side effect worth knowing about. Since validator votes won’t show up as transactions anymore, the raw transaction count reported on Solana activity dashboards will drop — even though the actual number of user payments and trades stays exactly the same. The Solana Foundation has told data providers to update their metrics to account for this. If they don’t, the numbers will look like activity fell off a cliff when it didn’t.

Transaction history services face a more specific technical requirement. They’ll need to keep competing candidate blocks separate until the network picks a final winner. If they don’t, transaction records could come out inaccurate. That’s a real migration burden for some infrastructure providers.

More context: Blockchain Association CEO Summer Mersinger Resigns After Senate Vote Halts Key Stablecoin Bill

Applications that simply send transactions or read account balances, though? They don’t need to change anything. The Foundation was pretty clear on that point.

Timeline Still Murky, Speed Claims Still Theoretical

Here’s the catch. The 150-millisecond figure isn’t from live network conditions — it’s from simulations. Real-world performance could vary once mainnet traffic, wallet processing delays, and exchange deposit checks all pile on. Those extra layers add time that simulations don’t always capture cleanly.

And there’s no firm launch date. None. Anza’s software schedule seems to hint that feature activations could pick back up around September 28, but that date hasn’t been named as Alpenglow’s official rollout window. It’s probably not wise to treat it as one.

So where does that leave things? Testing continues. Developers are stress-testing on devnet and testnet. The Solana Foundation and Anza are watching the results. Stakeholders are waiting.

The practical upside, if Alpenglow performs anywhere close to its targets on mainnet, would be felt in some pretty specific places. Faster finality means exchanges can release funds sooner after a trade settles. Merchants processing crypto payments could see confirmation times drop from something that feels sluggish to something that feels instant. For DeFi protocols where speed matters — liquidations, arbitrage windows, order routing — 150 milliseconds is a different world than 12.8 seconds.

Data Providers on Notice

The data reporting issue is worth sitting with for a second. Solana already gets scrutinized heavily for how it counts transactions, partly because validator votes historically inflated raw numbers. Alpenglow removes those votes from the count, which should actually make the data cleaner and more comparable to other networks. But the transition period is messy. If data providers are slow to adjust, you’ll see charts that look like Solana’s network activity cratered overnight. It didn’t — the methodology changed.

More context: Bitget Hit by $351.6 Million Hack Targeting Warm Wallets, Not Private Keys

The Foundation’s guidance is clear: adjust the metrics now, before mainnet, so the comparison baseline is set correctly.

It’s a small thing in the grand scheme of a consensus-layer overhaul. But in crypto, where narrative moves markets, a chart that looks wrong can do real damage fast.

Anza hasn’t specified exactly when it plans to push for a mainnet activation vote among validators. That vote still has to happen — and validators have to agree. No date, no guarantee. Testing on the developer and test networks continues.

Hub: Solana price, news, and analysis