Markets
BTC $60,555 +2.99% ETH $1,700 +2.20% BNB $553 +3.28% XRP $0.9211 +1.69% SOL $70.06 +1.10% LINK $8.77 +2.91% BTC $60,555 +2.99% ETH $1,700 +2.20% BNB $553 +3.28% XRP $0.9211 +1.69% SOL $70.06 +1.10% LINK $8.77 +2.91% TOTAL $2.07T BTC.D 58.8%

Altcoins

Solana’s 2027 Roadmap Doubles Compute for a Chain That Isn’t Full

Anza's plan adds capacity, cheaper token accounts and onchain AI models, while Solana's fee take suggests block space is not yet the binding constraint.

Anza used Breakpoint to publish a 2027 roadmap for Solana, and the network’s own summary of it runs to four lines: compute capacity doubles, token accounts get 80% cheaper in compute units, vector processing arrives so AI models can run fully onchain, and the protocol gets simpler.

Brennan Watt, who runs the validator client team spun out of Solana Labs, presented the plan on November 19. The post carries the details that matter. The account limit per transaction rises from 64 to 96, a 50% increase in the account footprint of a single transaction. Account data is read rather than copied, cutting the bytes moved per transaction. Both changes sit alongside Alpenglow, described as next with faster block distribution, and Constellation, aimed at clearer ordering rules and more competition among validators.

Taken together, the arithmetic compounds. An 80% cut in the compute cost of the most common account operation means the same block budget carries five times as many of those operations; doubling network capacity then doubles the budget. On the roadmap’s own numbers, a block that clears a given volume of token transfers today could clear several times that volume on the same fee schedule, before any change in bandwidth.

The bandwidth detail matters more than it reads. Reading account data instead of copying it reduces the bytes moved, and bytes rather than raw compute are what has historically produced congestion on Solana. Vector processing for models running onchain is a different kind of bet: it prices inference as a native workload rather than something bolted onto a general-purpose chain through a bridge or oracle.

The timing is the awkward part, because capacity is being added to a network whose blocks are not full. DefiLlama puts Solana at $317,052 in fees over the last 24 hours, $3.00 million over seven days and $13.81 million over 30 days. Ethereum recorded $236,218, $2.30 million and $20.15 million over the same windows.

Solana’s $40.73 billion market capitalisation is 20.1% of Ethereum’s $202.6 billion. Over 30 days it collected 68.6% as much in fees; over seven days, 129.9%; over 24 hours, 134.2%. On fee revenue the chain already behaves like a peer. On market value it does not.

That gap is why cheaper block space cuts both ways. Doubling compute and cutting the unit cost of token accounts lowers the price of the resource Solana sells. If demand grows faster than capacity, the network captures more activity at a lower unit price. If it does not, the same activity simply costs less and the fee line absorbs the difference.

The market-structure half of the roadmap is harder to price. Alpenglow and Constellation change how the chain decides what happens first and who stays in the validator set, which shows up in participation rather than benchmarks. A simpler protocol with fewer special cases is cheaper to change safely, an argument that tends to matter more after an outage than before one.

SOL traded at $68.80 at 08:20 UTC on Sunday, up 4.64% over 24 hours and down 2.81% over seven days — 76.5% below the $293.31 record set on January 19, 2025. The roadmap describes a chain building for load it has not yet attracted.

Share this article

Viber uses viber://forward?text= when the app is installed.