BNB Smart Chain post-quantum tests cut throughput 40% and raise P99 latency

BNB Smart Chain post-quantum planning is no longer a theoretical security exercise. BNB Chain has published a technical report mapping how BNB Smart Chain could migrate toward post-quantum cryptography, and the early results point to a blunt trade-off: stronger cryptography comes with much heavier data overhead.

That tension runs through the entire proposal. On one side, the report aims to prepare the network for future quantum threats. On the other, the tests show that much larger signatures can drag down throughput even when consensus behavior holds up relatively well.

The picture is unusually clear for a blockchain infrastructure study. BNB Chain is not just floating abstract ideas. It is laying out specific replacements for transaction signatures and validator vote aggregation, then showing what those choices do to block size, latency, and throughput.

BNB Smart Chain post-quantum upgrade path

What the report changes

The report centers on two major changes inside BNB Smart Chain.

First, transaction signatures would move from ECDSA to ML-DSA-44. Second, consensus vote aggregation would shift from BLS12-381 to pqSTARK aggregation.

That makes this BNB Smart Chain post-quantum roadmap notable for a simple reason: it touches both user transactions and validator coordination, the two places where cryptography directly shapes network performance.

The report also says existing wallets, SDKs, and RPCs would not need updates, and address formats would stay unchanged. That matters because compatibility is often where technically elegant upgrades become operationally messy. Here, the migration path appears designed to limit disruption at the interface level even as the underlying cryptography changes.

Why the timing matters

Post-quantum cryptography has become a bigger focus as blockchain developers think about long-term security, especially for systems still expected to be operating years from now. In this case, the report frames the challenge less as an immediate crisis and more as infrastructure preparation.

The strategic point is hard to miss: if a network waits too long to test quantum-resistant signatures, it risks discovering performance bottlenecks only after the migration becomes urgent. By publishing benchmark-style results now, BNB Chain is effectively putting the cost of quantum-resistant signatures on the table early.

How pqSTARK aggregation changes consensus votes

What replaces BLS12-381

On the consensus side, BNB Chain proposed replacing BLS12-381 aggregation with pqSTARK aggregation.

That proposal looks important because it suggests the consensus layer may be more adaptable than the transaction layer. The report’s own framing points in that direction: the hardest part was not consensus itself, but the swelling transaction and block sizes that came with quantum-resistant signatures.

In practical terms, that means validator coordination may be easier to optimize than user-facing transaction data growth. For blockchain networks, that is a meaningful distinction. If consensus remains comparatively efficient, developers can focus their attention on the part of the system that hurts scalability most.

How much compression improves

The most striking number in the report comes from validator signature compression. Six validator signatures totaling 14.5 KB compress into a single proof of about 340 bytes using pqSTARK aggregation.

That is a dramatic reduction, and it helps explain why the consensus change looks more manageable than the transaction signature change. Compression at that level gives the network a way to offset what would otherwise become a much larger validator messaging burden under post-quantum cryptography.

Why this matters: blockchains often struggle when security improvements increase bandwidth demands faster than the network can absorb them. pqSTARK aggregation appears aimed squarely at that problem, reducing signature bulk where validators need efficiency most.

Why larger signatures raise the performance cost

Transaction size growth

The biggest cost in the report sits at the transaction level. ML-DSA-44 signatures increase per-transaction signature size from 65 bytes under ECDSA to 2,420 bytes.

That jump is enormous, and it explains why the report treats data growth as the main challenge for any production rollout. In blockchain systems, bigger signatures do not just affect storage. They also affect block composition, bandwidth pressure, and how quickly information can move across geographically distributed nodes.

This is the core trade-off behind the BNB Smart Chain post-quantum effort. Quantum-resistant signatures may strengthen future-proofing, but they also make each transaction much heavier before any other optimization kicks in.

Throughput and finality impact

The tests showed the performance impact clearly.

• Native transfer throughput dropped by about 40%, while gas throughput fell by roughly 50% under cross-region conditions.
• Median finality stayed at two slots, but P99 finality latency rose to 11 slots because of block propagation delays.

Those numbers tell an important story. The network did not appear to lose its basic ability to finalize blocks at the median. However, at the tail end of performance, latency worsened, especially when nodes were spread across regions and had to move larger blocks over wider network links.

That distinction matters. A stable median finality of two slots suggests the consensus mechanism itself remained steady in testing. The rise in P99 finality latency to 11 slots points instead to networking stress from bigger blocks, not a collapse in protocol logic.

In other words, stronger cryptography did not break finality. It made the system heavier to carry.

What the BNB Smart Chain post-quantum trade-off means

For users and developers, this is the real takeaway. The report suggests that moving to quantum-resistant signatures on BNB Smart Chain is technically feasible, but the cost shows up in throughput and data load far more than in consensus stability.

That has broader consequences for network design. If the binding constraint becomes block bytes before gas limits are reached, then future optimization work may need to focus less on pure execution efficiency and more on how transaction data is encoded, transmitted, or compressed.

It also gives a clearer view of what post-quantum readiness actually means in blockchain infrastructure. It is not just about swapping one cryptographic primitive for another. It is about whether the surrounding system can absorb the size and bandwidth shock that comes with that swap.

And that may be the most revealing part of the report. BNB Chain’s proposed path suggests the consensus layer can be modernized with relatively strong compression through pqSTARK aggregation, while the transaction layer remains the harder battlefield. If quantum-resistant signatures are the future, the race may not be about who adopts them first, but who figures out how to carry all that extra weight without choking throughput.