Researchers Slash Quantum Computing Requirements for Bitcoin and Ethereum Attack

Highlights:

• Researchers made a key step in a quantum attack much more efficient against Bitcoin and Ethereum cryptography.
• The new benchmark dropped by 86.1%, and researchers also reduced the qubits and quantum operations required.
• Still, Bitcoin and Ethereum are not facing an immediate quantum attack, the researchers say.

Researchers have sharply reduced the estimated quantum computing resources needed for a key operation involved in attacking the cryptography used by Bitcoin and Ethereum. The new work focuses on making Shor’s algorithm more efficient against secp256k1, the elliptic curve used for transaction signatures across both networks.

The research paper, published on arXiv on September 9, shares the results of ECDSA.Fail, a research challenge involving more than 100 people and AI agents. The main goal was to find a way to make an important part of a quantum attack more efficient. Researchers focused on reducing the computing power needed for a process called elliptic-curve point addition, which is used when Shor’s algorithm tries to break elliptic-curve cryptography.

At the beginning of the challenge, researchers needed 2,715 logical qubits and nearly 4 million Toffoli gates to complete the process. By the July 26 cutoff, they had reduced those requirements to 1,151 qubits and about 1.3 million Toffoli gates. In simple terms, the researchers found a much more efficient way to perform the same quantum operation. The overall benchmark improved by 86.1%, while the number of qubits needed fell by 57.6% and the Toffoli gate count dropped by 67.2%.

A MAJOR quantum leap. 🔥

Researchers working with AI agents cut a benchmark for one step in a potential Bitcoin and Ethereum attack by over 50%, per CoinDesk.

Over 100 participants spent roughly eight weeks improving a calculation that could eventually help derive private keys… pic.twitter.com/eelVRw9HHy

— Coin Bureau (@coinbureau) September 10, 2026

New Circuit Beats Earlier Quantum Benchmarks

The researchers compared their result with earlier work from Google Quantum AI and other researchers. The best ECDSA.Fail circuit recorded a Q × T score more than 50% below Google’s reported point-addition benchmark. However, the paper stresses that differences in interfaces, assumptions and accounting methods prevent a direct claim that one approach fully outperforms another. 

The researchers also created another version of the circuit that works with the method used in Shor’s algorithm. It needed 1,162 logical qubits and about 1.68 million Toffoli gates. In tests using 100,000 random inputs, the circuit worked successfully about 99.8% of the time. The team kept improving the design after the July cutoff. One later version reduced the benchmark score to about 1.26 billion, while another brought the number of required logical qubits down to just 813.

Bitcoin and Ethereum Are Not Facing an Immediate Quantum Attack

The findings matter for Bitcoin and Ethereum because secp256k1 protects their transaction signatures. A sufficiently powerful cryptographically relevant quantum computer running Shor’s algorithm could, in principle, recover a private signing key from an exposed public key and use it to forge signatures. 

However, the researchers make clear that their work does not demonstrate a working attack on Bitcoin or Ethereum. The benchmark covers one important quantum-circuit component rather than a complete fault-tolerant implementation of Shor’s algorithm. As the paper explains, the results “refine the understanding of one logical-resource bottleneck rather than establish an immediate capability to break secp256k1.” 

The benchmark does not include physical error correction, hardware architecture, circuit depth, or other real-world costs. The next research step is to integrate the optimized point-addition circuit into a complete windowed implementation of Shor’s algorithm and measure the end-to-end resource requirements.