Jump Crypto’s Firedancer team plots upgrade to reduce transaction time to 100 milliseconds

Jump Crypto’s Firedancer team has presented a proposal, SIMD-0370, to eliminate Solana’s fixed per-block compute unit (CU) limits. The company said the initiative aims to strengthen the network’s performance and incentivize validators with suboptimal hardware to upgrade.

The Web3 infrastructure firm’s initiative will be implemented after the Alpenglow upgrade, which passed in a near-unanimous vote earlier this month. The Alpenglow upgrade will also be deployed on a testnet in December.

Alpenglow upgrade allows blocks to scale dynamically

Solana Research company, Anza, argued on Saturday that removing static block caps would allow slower validators to skip more complex blocks, leaving them for better-equipped validators to handle. The company added that it will create a performance flywheel, where block producers pack more transactions to earn more fees and drive overall network capacity.

The Firedancer team’s proposal aims to eliminate Solana’s current fixed block compute unit limit of 60 million CUs. Removing a fixed limit will allow the block size to scale based on how many transactions a validator can fit into a block.

The initiative builds on the previous discussions (SIMD-0286) to increase the network’s fixed block compute unit limit to 100 million CUs. Jump Crypto said it’s advocating for blocks to scale dynamically, limited only by the processing capabilities of the most performant validators on the network.

Solana’s Alpenglow upgrade contains the dynamic scaling mechanism, which is scheduled for testnet deployment in December 2025, with mainnet activation expected in Q1 2026.

Jito Labs’ CEO, Lucas Bruder, introduced the SIMD-0286 in May to increase the compute block limit to 100 million CU. Engineer Akhilesh Singhania said on GitHub that the upgrade may also create centralization risks despite being aimed at incentivizing validators to upgrade hardware to earn more fees.

The upgrade to Solana’s protocol introduces new consensus mechanisms (Votor and Rotor) that replace the traditional Tower BFT and Proof of History.

According to the announcement, Votor is expected to cut down on transaction finality, while Rotor aims to replace the proof-of-history timestamping system and reduce the time required for data transfers between validators. The upgrade aims to reduce transaction finality from approximately 12.8 seconds to roughly 100-150 milliseconds.

Solana plans to unlock new use cases requiring speed and cryptographic certainty 

The Solana Foundation said last month that reducing speeds to 150 milliseconds could help Solana realize Web2-level responsiveness with L1 finality, which would unlock new use cases that require both speed and cryptographic certainty. The company added that the compounding effect of the initiative in the Solana ecosystem is a financial infrastructure that operates at internet speed.

The Alpenglow upgrade also introduces a skip-vote mechanism, which enables slower validators to automatically abstain from voting on oversized blocks they cannot process in time without disrupting the chain.

Solana’s sudden surge in network activity has led to network outages in the past, prompting the need for additional upgrades to maintain the network’s stability. Jump Crypto revealed that the Solana Foundation will oversee the broader ecosystem development and governance, while Anza will spearhead the Alpenglow upgrade. 

Anza argued that a successful implementation of the protocol would be the biggest change to Solana’s network. The company also believes that SIM-0370 will position Solana to compete with the current internet infrastructure.

Jump Crypto’s Firedance development team proposal states that the current incentive for validator clients and program developers is broken. The team also argued that the capacity of the network is determined not by the capabilities of the hardware but by the arbitrary block compute unit limit.

Earlier this month, the company provided funding for Forward Industries’ $1.65 billion PIPE deal, alongside Galaxy Digital and Multicoin Capital.

If you're reading this, you’re already ahead. Stay there with our newsletter.