BNB Chain’s new BSC Post-Quantum Cryptography Migration Report shows that the chain has already tested ML-DSA-44 for post-quantum transaction signatures and used pqSTARK to aggregate consensus votes. The design worked with the current BSC stack. The trade-off was clear: much larger transaction data and weaker throughput in testing.
Signature growth pushed transaction size sharply higher
The report identifies data expansion as the main issue. After replacing ECDSA with ML-DSA-44, transaction signatures increased from 65 bytes to 2,420 bytes. Full transaction size rose from 110 bytes to about 2.5 KB. That change fed directly into larger blocks, which reached roughly 2 MB in the tests, and overall throughput dropped by about 40% to 50%.
Under cross-region conditions, TPS fell by around 40%. BNB Chain said the slower performance came from the harder job of propagating larger blocks across regions, not from a breakdown in the consensus model. Median finality still held at two slots, while slower outliers were linked to block movement delays.
Current addresses, RPCs, SDKs, and wallets remain compatible
BNB Chain said the migration path remains compatible with existing addresses, RPCs, SDKs, wallets, and transaction flows. If the approach reaches production, users and developers would not need to change basic account formats first. The team also wrote that post-quantum readiness is achievable on BSC today, while adding that quantum computers are not yet at a point where they can break production cryptography in real-world systems.
pqSTARK reduced validator-side overhead during tests
Consensus vote aggregation held up better than the transaction layer. According to the report, pqSTARK delivered about 43:1 compression, which helped keep validator overhead manageable in the test environment. That points to the transaction data layer, rather than vote aggregation, as the heavier pressure point in the current design.
The migration also leaves parts of BSC’s cryptographic stack untouched for now. Peer-to-peer handshakes and KZG commitments are outside the current scope. The report says P2P migration would require ML-KEM, while replacing KZG would need coordination across the broader Ethereum ecosystem.
Scaling work still stands between testing and production
The result adds a new constraint to BNB Chain’s broader performance roadmap. Related reports have said the network is targeting sub-150 millisecond finality and more than 20,000 TPS for complex transactions by 2026. This test shows BSC can adopt ML-DSA-44 and pqSTARK, but much larger signatures make those throughput goals harder to reach unless data handling and network scaling improve.
BNB Chain’s report leaves little ambiguity on that point. The chain can move toward quantum-resistant security, but network and data-layer scaling remain the main barriers before any production deployment.