AdvertiseFee Deal Desk

Solana

Anza's 2027 Solana (SOL) Roadmap Doubles Compute Capacity to 500M CU

Anza CEO Brennan Watt laid out a 2027 Solana roadmap at Breakpoint, doubling compute capacity to 500M CU per second and cutting token account costs by 80%.

Be a creator
November 19, 2026, 02:50 PM UTC4 min read
AI SummaryAI
  • Anza CEO Brennan Watt presented a 2027 Solana roadmap at Breakpoint on November 19
  • Alpenglow went live on Solana mainnet on October 21, cutting finality from 12.8 seconds to about 150 milliseconds
  • Solana compute capacity is set to double from 250 million to 500 million compute units per second
  • Token account operations will use about 80% fewer compute units under the 2027 performance plan

Anza's 2027 Plan From the Breakpoint Stage

Validators and client teams now have the outline they will build against through 2027. On Thursday at the 2026 Breakpoint conference, Anza CEO Brennan Watt laid out next year's roadmap for Solana (SOL) across four tracks: resilience, market structure, performance and expressivity. The plan doubles network compute capacity and cuts the compute cost of token account operations by roughly 80%. Watt stayed off market matters, and the Solana price did not feature anywhere in the material presented.

The announcement follows a 2026 that Watt framed as the network's most consequential build year: Solana migrated its consensus to Alpenglow, live on mainnet since October 21, 2026, which replaced the TowerBFT voting mechanism with Votor, a faster path to agreement. Transaction finality fell from about 12.8 seconds to roughly 150 milliseconds, while block times were halved from 400 milliseconds to 200 milliseconds with capacity doubled alongside. The network also crossed to version 1.0, retiring the “mainnet beta” label it had carried since launch, a change Watt presented as recognition of progress on performance, resilience and developer experience. The official Solana (SOL) account compressed the four-track plan into a single summary post on November 19. On resilience, Watt told the audience the team will “simplify the protocol, replacing legacy abstractions and interfaces” and push more execution on chain, changes he said allow the protocol to evolve and ship faster. Validators, whose staking weight ultimately decides which protocol changes activate, sit at the center of the market structure track detailed below. Breakpoint has been the venue for several structural moves this quarter, including the Solana Foundation's majority stake in CoinGecko, a platform with 18 million users. Storage economics also moved this year: the cost of storing data on Solana fell 27%, and Watt said a further 7x reduction arrives within roughly two weeks.

500 Million CU and an Undated Relay

The performance track carries the largest figures. Network compute capacity is set to double from 250 million to 500 million compute units (CU) per second, the unit that meters how much processing each transaction may consume, lifting how many transactions the chain can settle each second. Watt's 2026 capacity doubling came with the block time cut, and the 2027 target repeats that pattern. Token account operations, the touchpoint behind most transfers and program interactions, are slated to use about 80% fewer CU, a reduction Watt presented alongside the capacity doubling.

Expressivity work adds vector processing to the virtual machine so compute-heavy dapp workloads and models can run fully on chain, lifts the account limit per transaction from 64 to 96, and switches to reading account data directly instead of copying it, which cuts how much data moves per transaction. The current 64-account ceiling constrains transactions that touch many programs at once. That direction extends usage the Solana (SOL) Foundation has already reported on, where 76% of OpenRouter token use now comes from AI agents. The market structure track extends Alpenglow itself. Alpenglow.Next shortens the wait between block-producing leaders so blocks distribute faster, while Rotor, a single relay layer, is slated to replace Turbine, the protocol that spreads block data across the network, with no activation window announced. Alpenglow's rollout runs in two phases: Votor, its voting protocol, is already live, and Rotor is the later piece. Constellation introduces clearer ordering rules for transaction inclusion and aims to widen competition among validators rather than leaving ordering defaults implicit. That validator set is already an institutional touchpoint: ICE has weighed a validator role since remarks at Breakpoint on Nov 16. Resilience rounds the plan out with fewer special cases, more on-chain execution and modernized wire formats, the same simplification Watt described on stage.

For the parties downstream, the load-bearing items sit in two places. The compute doubling and the 80% CU cut change the cost base for high-throughput DeFi and payments applications directly, and the ordering rules under Constellation touch the same plumbing that carries regulated flows such as tokenized equities, a market that reached $1.22 billion on the Solana network. Neither deliverable carries a ship date yet: Rotor in particular has no activation window, so the nearest dated commitment in the whole plan is the 7x storage cost reduction Watt set for within roughly two weeks. Until client teams ship implementations and validators signal readiness, the 2027 targets remain a schedule rather than a changelog.

Readers tracking the market in real time can follow live spot and futures prices on Bybit.

Primary sources

COINOTAG's editorial and research desk.

AI-Assisted

AI-generated, AI-reviewed, under COINOTAG editorial oversight.