BNB Chain has released its BSC Post-Quantum Cryptography Migration Report, documenting a successful test of post-quantum cryptographic primitives while acknowledging that cross-region throughput fell by approximately 40% and that network and data-layer scaling remain unsolved before production deployment.
Key Takeaways
ML-DSA-44 replaces ECDSA; signature size grows from 65 bytes to 2,420 bytes, +37x.
Block size at 2,000 TPS grows from ~130 KB to ~2 MB, +18x, driven by signature size.
Cross-region TPS falls approximately 40% in testing.
Existing addresses, RPCs, SDKs, and wallets remain compatible.
What the upgrade replaces and why
The existing ECDSA algorithm using the secp256k1 curve is computationally efficient but theoretically vulnerable to sufficiently powerful quantum computers. Based on publicly documented cryptographic research, that vulnerability stems from quantum algorithms capable of attacking the mathematical foundations elliptic curve schemes depend on. The upgrade replaces it with ML-DSA-44, a lattice-based scheme standardized under NIST FIPS 204. Consensus vote aggregation moves from BLS aggregate signatures to pqSTARK, where a single proof covers all validator signatures. Existing addresses, RPCs, SDKs, and wallets remain compatible.
Where the overhead accumulates and why it costs 40%
Public key size grows 20x, according to Wu Blockchain report, from 64 bytes to 1,312 bytes. Signature size grows 37x, from 65 bytes to 2,420 bytes. The signature is the dominant burden: every transaction carries one, and the table confirms the +18x block size increase from approximately 130 KB to approximately 2 MB is driven directly by per-transaction signature size.
The 40% TPS decline is not a standalone result: it is the operational endpoint of a causal chain that begins with a 37x signature size increase, produces an 18x block size expansion, and ends with cross-region propagation too slow to sustain current throughput. The cryptographic layer has been solved. The propagation layer has not.
Most encouraging number
The consensus layer tells a different story from the transaction layer: pqSTARK replaces BLS aggregation with a single proof covering all validator signatures, growing only 3.5x, which means the upgrade’s efficiency problem is concentrated entirely at the per-transaction level, not at the network’s coordination layer. Aggregated proof size moves from 96 bytes to approximately 340 bytes. Because the consensus layer is already efficient under the new scheme, the remaining engineering work is concentrated at one point: reduce per-transaction signature overhead without compromising quantum resistance.
BSC Post-Quantum Upgrade Passes Test, but TPS Falls 40%
BNB Chain released its BSC Post-Quantum Cryptography Migration Report, saying BSC has tested a post-quantum upgrade using ML-DSA-44 for transaction signatures and pqSTARK for consensus vote aggregation. The design remains… pic.twitter.com/WKDKIwBanX
— Wu Blockchain (@WuBlockchain) May 19, 2026
Backward compatibility with existing addresses, RPCs, SDKs, and wallets removes the migration cost that has historically made cryptographic upgrades prohibitive for live networks, but it does not remove the throughput cost that will determine whether this design reaches production. A 40% TPS reduction on a network that competes on speed and cost is not deployable at current performance.
If a subsequent test report shows cross-region TPS decline below 15% with ML-DSA-44 intact, the upgrade is approaching production viability. If the gap remains above 30% after further optimization, the data-layer scaling problem will require a more fundamental architectural change than compression or caching alone can provide.
