Researchers have lowered the estimated computational requirements for a potential quantum attack on Bitcoin and Ethereum by more than half, highlighting how improvements in algorithms could bring such attacks closer even without a major breakthrough in quantum hardware.

A paper shared with CoinDesk details a new quantum circuit developed by researchers from the Ethereum Foundation, Theta Labs, StarkWare and other organizations. The circuit uses 1,151 logical qubits and roughly 1.3 million Toffoli gates, both of which help measure the resources required to perform the computation.

Logical qubits indicate the approximate size of the quantum machine, while Toffoli gates measure the amount of computational work involved. The researchers’ main design produced a score of about 1.5 billion, compared with approximately 3 billion in Google Quantum AI’s March benchmark.

That represents a reduction of more than 50%. However, the researchers cautioned that the two results are not directly equivalent because the designs use different interfaces and accounting methods.

A second design intended to more closely replicate how the calculation would function within Shor’s algorithm produced a score of around 1.96 billion, which also remained below Google’s earlier figure.

Researchers target point addition

The work focused on optimizing point addition, a calculation repeatedly performed by Shor’s algorithm.

The significance for Bitcoin and Ethereum comes from the algorithm’s potential ability to attack the cryptography protecting exposed public keys. With a sufficiently powerful quantum computer, an attacker could theoretically derive a private key from a public key and then use it to authorize transactions.

Both Bitcoin and Ethereum use secp256k1, the elliptic-curve cryptographic system examined by the researchers.

More than 100 participants contributed to the research over approximately eight weeks through ECDSA.Fail, an open challenge launched by Eigen Labs. The effort generated more than 400 accepted submissions, allowing successful improvements to serve as starting points for further optimization.

AI coding agents played a major role in implementation, repeated testing and smaller improvements. Human researchers generally selected the research directions and introduced larger changes to the designs.

The researchers did not attempt to determine how much of the final progress came specifically from humans or AI.

The findings nevertheless show that improvements in software and circuit design could shorten the quantum-security timeline independently of advances in physical hardware. If the required computation becomes more efficient, a smaller quantum system could potentially perform the same task.

Current quantum computers remain incapable of the attack

Despite the substantial reduction in computational requirements, the research does not mean Bitcoin or Ethereum can currently be compromised by quantum computers.

The circuit represents only one important calculation within a potential Shor-based attack. It does not include physical error correction, the complete Shor algorithm or several hardware-specific costs associated with running the computation on an actual quantum machine.

No existing quantum computer can use these results to break Bitcoin or Ethereum.

The researchers have also continued improving the designs since the paper’s July cutoff. One later version reduced the score to about 1.26 billion. Another design brought the machine requirement down to 813 logical qubits, although it required considerably more computation.

Jieyi Long, the paper’s lead author and CTO of Theta Labs, told CoinDesk that the results should not be interpreted as evidence of an imminent quantum attack. The urgency instead comes from the time required to develop and deploy protections, which cannot be implemented retroactively.

The research arrives as the U.S. government is also supporting development of more advanced quantum hardware. The Commerce Department finalized CHIPS Act awards of up to $100 million each for Rigetti, D-Wave and Quantinuum, taking minority stakes in the three companies. The funding is aimed at developing larger fault-tolerant quantum machines.

For the crypto industry, the latest results serve as another warning that the quantum threat is not determined solely by hardware progress. As researchers continue reducing the computational workload, Bitcoin and Ethereum developers may need to prepare quantum-resistant upgrades well before a machine capable of executing a full attack actually exists.