Alpenglow activated at epoch 1064, officially retiring TowerBFT, which Solana had used for more than six years. The official claim is that it cuts finality from about 12.8 seconds to 150 milliseconds—a 98.8% reduction—but that figure isn’t the real story.
What changes under the hood is the voting mechanism. In the TowerBFT era, validators had to submit votes as on-chain transactions, and a block needed to accumulate votes across 32 slots to become finalized. Votor lets validators exchange votes directly and aggregate them into certificates: one round to finalize with 80% of stake voting in favor, or two rounds with 60%. The key side effect: votes are no longer recorded on-chain, even though vote transactions previously accounted for about 56% of Solana’s recorded transactions. In other words, TPS and transaction counts on various dashboards will “plunge” after the upgrade—but that doesn’t mean users have left.
That’s where opinions diverge. Bulls say 150-millisecond finality makes Solana competitive with centralized matching engines for payments, exchange deposits, and cross-chain bridges. Bears say finality is about latency, not throughput; fees won’t fall, and the marketing numbers will soon come back to bite. There’s also a third point that’s been overlooked: each validator must keep at least 0.8 SOL in its vote account to pay the entry fee for each epoch (which is burned). This creates a new barrier to entry that disadvantages smaller validators.
I’m inclined to think Alpenglow’s biggest significance isn’t speed, but the idea of “decoupling consensus costs from user transactions” itself—the network will no longer make user transaction volume pay for consensus. But to verify that, we need to look at the actual distribution of finality times after activation, not the target values in the white paper. The switchover is happening in stages today: the first 5,000 slots (about 17–18 minutes) will still run TowerBFT, with a pause of about 400 milliseconds during the handoff.
Here’s the question: TPS cut in half, finality about 85 times faster—which would you believe first? For this wave, should we focus on the narrative or on actual deposit speeds? $SOL
#SolanaWillActivateAlpenglowUpgradeToday
What changes under the hood is the voting mechanism. In the TowerBFT era, validators had to submit votes as on-chain transactions, and a block needed to accumulate votes across 32 slots to become finalized. Votor lets validators exchange votes directly and aggregate them into certificates: one round to finalize with 80% of stake voting in favor, or two rounds with 60%. The key side effect: votes are no longer recorded on-chain, even though vote transactions previously accounted for about 56% of Solana’s recorded transactions. In other words, TPS and transaction counts on various dashboards will “plunge” after the upgrade—but that doesn’t mean users have left.
That’s where opinions diverge. Bulls say 150-millisecond finality makes Solana competitive with centralized matching engines for payments, exchange deposits, and cross-chain bridges. Bears say finality is about latency, not throughput; fees won’t fall, and the marketing numbers will soon come back to bite. There’s also a third point that’s been overlooked: each validator must keep at least 0.8 SOL in its vote account to pay the entry fee for each epoch (which is burned). This creates a new barrier to entry that disadvantages smaller validators.
I’m inclined to think Alpenglow’s biggest significance isn’t speed, but the idea of “decoupling consensus costs from user transactions” itself—the network will no longer make user transaction volume pay for consensus. But to verify that, we need to look at the actual distribution of finality times after activation, not the target values in the white paper. The switchover is happening in stages today: the first 5,000 slots (about 17–18 minutes) will still run TowerBFT, with a pause of about 400 milliseconds during the handoff.
Here’s the question: TPS cut in half, finality about 85 times faster—which would you believe first? For this wave, should we focus on the narrative or on actual deposit speeds? $SOL
#SolanaWillActivateAlpenglowUpgradeToday