BNB Smart Chain prepares for validator upgrades by September 15

BNB Smart Chain validators began rolling out a significant capacity upgrade on September 14, with most expected to complete the transition by September 15. The core change: raising the target block gas limit to 70 million, a move designed to keep pace with a network already running at more than 95% utilization.

What’s actually changing

The upgrade is a direct follow-on to the Pasteur hard fork, which activated on August 25, 2026. That fork introduced BEP-675, a new block-building architecture that lets builders submit pre-executed blocks through a system called BidBlock V2.

Early mainnet data suggests the new approach is working. BidBlock V2 blocks have averaged a 28% increase in gas per block compared to the prior methodology, with peak (P99) blocks hitting 43.8 million gas. BidBlock V2 now manages roughly 98% of recent blocks on the network.

The on-chain gas limit has already crossed 55 million. But with utilization consistently above 95%, there’s little room left before users start feeling congestion in the form of slower confirmations and rising fees. Hence the push to 70 million.

One technical detail worth noting: validators aren’t simply plugging in a new hard-coded number. The 70 million target must be derived dynamically, meaning each validator calculates and adjusts the figure based on real-time conditions rather than a static setting. The framework also requires validators to refine their block-packing and simulation strategies to handle the larger blocks efficiently.

The Pasteur hard fork context

The Pasteur upgrade introduced BEP-675, which restructured how blocks get built by separating the roles of block proposers and block builders. BSC’s version, implemented through BidBlock V2, allows specialized builders to construct blocks offline and submit them pre-executed. This reduces redundant computation at the validator level and lets builders pack transactions more densely. The 28% average gas improvement per block is the direct payoff of that architectural choice.

Where this is heading

The 70 million gas limit target is not the finish line. BSC developers have outlined a sequential scaling plan that envisions pushing toward 80 million and potentially 90 million gas per block. Each step will depend on observed performance data from the mainnet.

There’s a risk dimension too. Larger blocks mean more data for validators to process and propagate. If some validators struggle with the increased load, it could lead to missed blocks or centralization pressure as only well-resourced operators keep up. The dynamic gas limit approach partially addresses this by allowing gradual adjustment.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.