Researchers from Theta Labs, StarkWare, and other organizations have presented a new estimate of the resources needed for a potential quantum attack on the cryptography of Bitcoin and Ethereum. Their developed scheme uses 1151 logical qubits and about 1.3 million Toffoli gates. Its estimate amounted to approximately 1.5 billion operations, which is more than 50% less than the figure of about 3 billion presented by the Google Quantum AI team in March. However, the authors note that direct comparison of results is limited due to differences in interfaces and counting methods. Another version of the scheme, more accurately corresponding to the practical implementation of Shor's algorithm, received an estimate of about 1.96 billion operations. Researchers optimized the point addition operation, which is repeatedly used when executing the algorithm.
Shor's algorithm theoretically allows the computation of a private key using its associated public key. With a sufficiently powerful quantum computer, this could enable the creation of digital signatures on behalf of cryptocurrency address owners. The research concerns the secp256k1 cryptography used by Bitcoin and Ethereum. More than 100 people participated in the work, who, together with AI agents, spent about two months optimizing the scheme as part of the open ECDSA.Fail trial created by Eigen Labs. In total, more than 400 decisions were made. AI was used for implementing proposals, retesting, and minor optimizations, while participants determined the main research directions and made more significant architectural changes. The authors did not separate the contributions of humans and AI in the final improvement.
The results obtained do not mean that Bitcoin or Ethereum can already be attacked by quantum computers. The research covers only one of the key computational parts of the attack and does not take into account physical error correction, the full Shor's algorithm, and a number of costs necessary to run it on a real device. Later versions of developments reduced the estimate to about 1.26 billion operations, and another approach reduced the required machine size to 813 logical qubits but required significantly more computations. The lead author of the work and CTO of Theta Labs, Jieyi Long, stated that there is no immediate threat yet, but transitioning to quantum attack protection may take years. Meanwhile, the U.S. Department of Commerce has completed the distribution of CHIPS Act grants of up to $100 million for Rigetti, D-Wave, and Quantinuum. The funds are intended for the development of larger fault-tolerant quantum systems.