Anza sets four engineering priorities for 2027

Anza, the core engineering company behind Solana's Agave validator client, has outlined a 2027 roadmap covering performance, expressivity, market structure and resilience. Solana summarized the plan on November 19 after Anza chief executive Brennan Watt presented it at the Breakpoint conference in London. The headline goals are to double network compute capacity and reduce the compute units consumed by token-account operations by 80%. Crypto News Flash reported that the capacity target would lift the network from roughly 250 million to 500 million compute units per second by the end of 2027. These are engineering targets, not changes already active on mainnet.

Cheaper token operations could free block space

Compute units measure the execution work required by transactions and programs. Reducing the cost of common token-account instructions could leave more of each block's compute budget available for other activity. That does not automatically mean an 80% cut in user fees or a matching increase in transactions per second, because transactions use different resources and fees also reflect signatures and priority bidding. Solana's official upgrade tracker shows that the network already raised its block limit from 60 million to 100 million compute units earlier in 2026. The roadmap proposes another capacity expansion alongside software changes intended to make frequently used token operations lighter.

Related reporting: Solana Activates Alpenglow Consensus on Mainnet

Transactions would gain more room and new math

The expressivity program includes vector processing for computational models running fully onchain, direct access to account data instead of repeated copying and a higher account limit per transaction. Solana's tracker lists the account-limit change under Agave 4.5, expected in December 2026, with version-1 transactions rising from 64 named accounts to 96. It remains marked in development. The same tracker lists account-data direct mapping as pending feature activation without a scheduled date. CryptoPotato reported that Watt framed the broader work around more complex financial applications, including options pricing and margin models, but the roadmap provides no published benchmark showing which models would become economical.

Rotor and Constellation address network coordination

Anza also plans changes to how blocks and transactions move through the network. Rotor is intended to replace Solana's existing Turbine block-distribution mechanism with a flatter design aimed at reducing latency. Constellation would introduce clearer transaction-ordering rules and competition among validators for inclusion, according to the Solana summary and independent reports. Those changes follow the October activation of Alpenglow, which replaced the previous consensus process and targets roughly 150-millisecond finality. The official tracker marks Alpenglow as live, but Rotor and Constellation are roadmap projects rather than completed upgrades.

Delivery will be measured feature by feature

The resilience objective is to simplify the protocol and reduce special cases so future changes can be implemented more safely. It is an engineering direction, not a guarantee against bugs, congestion or validator interruptions. Anza has not published one binding activation date for the complete roadmap, and individual changes will still move through development, client releases, validator adoption and network activation. The clearest progress checks are therefore the upgrade tracker, implementation proposals and mainnet status for each feature. Doubling compute capacity, cutting token-account CU use and changing block ordering would be material if delivered, but the November announcement establishes a plan rather than proof of those results.

Sources

AI-generated editorial image; not a photograph of the reported event. Prepared with AI assistance and source verification.