Solana completed its Alpenglow consensus upgrade on October 21, 2026, replacing the TowerBFT system that had governed transaction finality since the blockchain launched in 2020. The upgrade targets finality of approximately 150 milliseconds, compared with the previous benchmark of 12.8 seconds, representing a potential improvement of more than 98%.
However, an early reading from the Solana Foundation’s Alpenglow dashboard showed observed finality of 202 milliseconds, approximately 35% above the stated target.
The figure represents a substantial improvement over Solana’s historical finality performance but does not establish that the network has consistently achieved the advertised 150ms threshold. Importantly, the 202ms measurement is an individual observation rather than a sustained average.
The distinction matters because Alpenglow is intended to improve settlement certainty for cryptocurrency exchanges, payment providers, decentralized trading platforms and other applications that depend on rapid transaction finalization.
Alpenglow Replaces TowerBFT as Mainnet Transition Completes
According to the Solana Foundation, Alpenglow activated during epoch 1064 on October 21, which began at approximately 13:36 UTC. The consensus transition occurred at slot 459,653,000, following a genesis block at slot 459,652,999.
The genesis certificate received votes from 446 of Solana’s 654 validators, confirming sufficient participation to establish the new consensus mechanism. Alpenglow introduces Votor, a voting protocol that fundamentally changes how validators reach agreement. Previously, TowerBFT required validators to submit votes as on-chain transactions and accumulate confirmations across multiple slots.
Under Votor, validators exchange signed votes directly and aggregate them into certificates, allowing blocks to reach finality after one or two voting rounds. The modification also eliminates validator vote transactions from blocks, reducing the amount of network capacity consumed by consensus-related activity.
Consequently, reported transactions per second may decline significantly following activation without a corresponding reduction in actual user activity. Historical throughput figures that included validator votes are therefore not directly comparable with post-upgrade measurements.
Alpenglow also separates transaction finality from block production. Solana independently reduced its target slot time from approximately 400 milliseconds to 200 milliseconds during October, while the new consensus mechanism seeks to finalize transactions in roughly 150 milliseconds.
What the 202ms Reading Means for Solana’s Next Phase
The initial 202ms observation indicates that Alpenglow can deliver finality considerably faster than Solana’s previous consensus system, although the network’s sustained performance remains to be established.
Relative to the historical 12.8-second benchmark, 202ms represents an approximately 98.4% reduction in finality latency. Nevertheless, the comparison requires context because Solana’s previous finality times also improved as slot durations declined.
For exchanges and payment processors, faster finality could shorten the time required to credit deposits, confirm settlements and release funds. However, actual application-level processing times will also depend on infrastructure, network conditions and operational requirements.
The upgrade’s next major component, Rotor, remains under development. Rotor is intended to replace Solana’s existing Turbine block-distribution system, potentially improving how transaction data propagates across the validator network.
For now, the critical performance indicators will be sustained finality latency, network stability, validator participation and transaction processing under different activity levels.
The 202ms reading is therefore neither evidence of a failed upgrade nor confirmation that Solana has achieved consistent 150ms finality. It establishes an early post-activation reference point against which subsequent mainnet performance can be evaluated.
