1 hr ago3 min readCrypto researchers cut Bitcoin and Ethereum quantum attack estimate by 50%. (dujin yun/Pixabay)SummaryResearchers cut the estimated resources needed for a key step in a future quantum attack on Bitcoin and Ethereum by more than half compared with a Google benchmark published in March.The work optimized a calculation used in Shor’s algorithm, which could allow a sufficiently powerful quantum computer to derive private keys from exposed public keys and forge transactions.No existing quantum computer can use the research to break either network, but the findings underscore the need to begin adopting quantum-resistant protections because upgrades could take years.The estimated work needed for a key step in a future quantum attack on Bitcoin and Ethereum has fallen by more than half from a benchmark Google published in March.A paper shared with CoinDesk by researchers from the Ethereum Foundation, Theta Labs, StarkWare and several other organizations describes a quantum circuit using 1,151 logical qubits, or such a machine’s smallest working units, and about 1.3 million Toffoli gates, a measure of how much work it has to do.Think of the two numbers as machine size and workload, where fewer qubits means a smaller machine, while fewer gates means less work for that machine to do.The new circuit scores about 1.5 billion, more than 50% below the roughly 3 billion score reported by Google Quantum AI in March. The paper cautioned that the comparison is not perfectly like-for-like, however, because the designs use different interfaces and accounting rules.Read More: Here's what 'cracking' bitcoin in 9 minutes by quantum computers actually meansA second version adapted more closely to how Shor's algorithm would actually use the calculation scored about 1.96 billion, still below Google's reported figures.The researchers optimized point addition, a calculation performed repeatedly inside Shor's algorithm.Shor is the quantum algorithm that could eventually turn an exposed public key into the private key behind it. A powerful enough machine could then sign transactions as the wallet's owner. Bitcoin and Ethereum both use secp256k1, the cryptography targeted by the research.More than 100 participants, working alongside AI coding agents, spent roughly eight weeks attacking the problem through ECDSA.Fail, an open challenge created by Eigen Labs. They produced more than 400 accepted submissions, with successful improvements becoming a new starting point for others to build on.AI agents were used heavily for implementation, repeated testing and smaller optimizations, while humans tended to choose research directions and make larger changes to the design. The paper does not attempt to separate exactly how much of the improvement came from humans versus AI.This means quantum hardware does not have to improve for the attack to get closer. Researchers can also reduce how much machine the attack needs.Read More: Bitcoin and Ethereum race quantum clock as U.S. backs $300 million hardware pushThe researchers did not optimize an entire attack. Their circuit covers one of its major calculations and leaves out physical error correction, the full Shor calculation and some of the hardware-specific costs of running it on a real quantum computer. No existing machine can use the result to break Bitcoin or Ethereum today.The number has also kept falling, as the paper's headline result reflects a July cutoff. A later design included in the research pushed the score down again to about 1.26 billion, while another got the machine-size requirement down to 813 logical qubits at the cost of far more computation."None of this is urgent because an attack is imminent," lead author and Theta Labs CTO Jieyi Long told CoinDesk. "It is urgent because the remedy takes years and cannot be applied retroactively."The paper comes as the U.S. Commerce Department this week finalized CHIPS Act awards worth up to $100 million each for Rigetti, D-Wave and Quantinuum, taking minority stakes in all three, money aimed at building the larger fault-tolerant machines an attack would eventually run on. 12345678910