Solana Could Drop Block Limits as Firedancer Challenges Compute Unit Cap

Jump Crypto’s Firedancer team has introduced SIMD-0370, a proposal that could reshape how Solana processes transactions.

The independent validator client wants to remove the network’s fixed compute unit (CU) block limit, arguing that validator performance should determine capacity rather than an arbitrary ceiling.

Solana Developers Split Over Plan to Replace Block Cap

The proposal builds on Alpenglow, a forthcoming network upgrade that will reduce block finality from 12.8 seconds to as little as 100–150 milliseconds.

Alpenglow is expected to unlock far greater efficiency for the blockchain network by reducing congestion and eliminating redundant gossip messaging.

Firedancer contends that in such an environment, keeping Solana’s block capacity capped between 60 million and 100 million compute units, as mandated by SIMD-0286, is an unnecessary constraint.

Currently, every validator faces the same ceiling regardless of hardware. This structure, the team argues, prevents stronger machines from processing larger blocks and creates uneven incentives for developers and operators.

“The current incentive structure for validator clients and program developers is broken. The capacity of the network is determined not by the capabilities of the hardware but by the arbitrary block compute unit limit,” the team argued.

However, that would change with Firedancer’s SIMD-0370 proposal.

Under this proposal, block producers could pack as many transactions as their systems can handle.

Validators unable to process those blocks in time would simply skip them, while the chain would continue without disruption.

Firedancer maintains that this approach aligns network capacity with market demand. It creates a dynamic system where throughput scales up or down based on usage rather than manual updates.

The proposal also introduces more substantial incentives for competition.

Block producers who optimize their performance could include more transactions per block, thereby earning higher rewards.

In turn, slower validator clients must improve their setups to avoid falling behind and missing out on revenue.

Firedancer expects this to spark a “flywheel effect” in which consistent performance improvements raise the baseline capacity of the entire validator set.

“The net result is that the capacity of the network is governed by market forces – if demand is there, the capacity of thenetwork will increase to meet it,” the developers argued.

Still, not all developers are convinced about the plan.

Roger Wattenhoffer, head of research at Anza, warned that removing the block limit could introduce technical risks and foster centralization.

However, he noted that these problems could be solved.

“If the speed advances during an epoch, we might fall below the 60/80 thresholds, at which point we only get skips, and we basically have to enter Alpenglow’s disaster scenarios,” the researcher stated.

Similarly, system engineer Akhilesh Singhania warned that large operators scaling into more expensive hardware could price out smaller validators.

He cautioned that this shift might concentrate the network in fewer hands.