This is a segment of the LightSpeed Newsletter. Subscribe to read the full edition.
For all speed and throughput, Solana has always been working on the fact that the technology that powers its performance also limits its possibilities.
The proven history, a cryptographic clock that helps distinguish Solana from other blockchains, gave the network a unique way to order events before consensus. However, the consensus itself was never resolved.
Instead, the POH had to be paired with Tower BFT, a Byzantine fault resistance mechanism layered over Solana’s blockstream. It worked, but it was complicated, difficult to optimize and was increasingly overtaken by the ambitions of the network and its builders.
Last week, Solana announced a proposal in favor of Alpenglow, both in the proof of history and in the Tower BFT.
Alpenglow is about rethinking the structure of time, consensus and data flow across the network. Alpenglow, developed by Anza, a core engineering team spun from Solana Labs, completely prevents innovations inherent in proof of history.
It introduces a streamlined consensus framework that disrupts sequence, voting, and block finality into a single integrated design.
The big, flashy headline here is that we are looking at global finality in under 150 milliseconds.
The problem isn’t that Poh is flawed, just as much as they had to pair it with a layer of adjustment to complete the block. The POH sequences events with verifiable delay capabilities and generates a ledger of time that the validator can use to agree to the order.
However, the sequence still had to be voted and still had to be confirmed and confirmed. And in a world where Solana is already pushing the 400ms block, finality (not production speed) has become a major bad bottleneck for the network.
Tower BFT took 12.8 seconds to complete the block. To hide that delay, Solana introduced a “optimistic confirmation” and gave Daps a probabilistic guarantee of finality. But that wasn’t enough. Finality really, really needs to match production.
To that end, Alpenglow introduces two core components: voting and rotors.
Voting is the new final engine, allowing consensus to be reached in one or two voting rounds. If 80% of the stock is online, you can achieve finality in one round. If only 60% responds, it returns to 1 second. If both passes run simultaneously and complete the first victory.
Meanwhile, rotors re-adapt how data spreads through the network. Built as an improvement to Solana’s turbine protocol, rotors replace the turbine layered tree with a single flat relay structure. It leverages erasure coding and deterministic relay allocation to broadcast data with minimum hops and maximum bandwidth.
It’s so fast that the speed of light – literal geographical delay between nodes – becomes the main limit. If you’re like me, you’re “1.21 Jigawatts?!” – ing and “great scott!” – I removed the furniture the first time I read it.
The result is a protocol that is as instantaneously final as everyone dreamed of. In a simulated testnet using Zurich as the reference node, 65% of Solana’s interest reached notarization in less than 50 minutes. Perfect finality landed at a median as low as 100ms at times.
This is not a potential fantasy. There is no probabilistic intermediate foundation. This is a true and verifiable cryptographic consensus faster than most Web2 APIs.
And even in disadvantaged conditions, Alpenglow remains performance. Its “20+20” resilience model can still achieve consensus while enduring hostile 20% and 20% of non-responsive stocks.
