Wall Street Divided on Quantum Readiness and Bitcoin Risk

Wall Street firms are taking different approaches to quantum computing after internal tests and public research produced mixed signals about near-term utility and cryptographic risk.

Goldman Sachs hired a small group of scientists three years ago and ran a pilot with Amazon to test whether quantum algorithms could improve portfolio optimization. The bank’s in-house work found the target problem would need about 8 million logical qubits and would take an impractical length of time on current hardware. Goldman later cut most of the team during a broader cost-reduction program. Current quantum machines contain fewer than 100 physical qubits.

JPMorgan Chase retained a team of more than 50 physicists, computer scientists and mathematicians working on quantum applications such as optimization, machine learning and cryptography. The bank continues internal research while monitoring improvements in hardware and algorithms.

Quantum bits, or qubits, differ from classical bits because they can exist in combinations of states before measurement. Properties called superposition and entanglement let quantum systems perform some calculations differently than classical computers. Researchers say potential applications include detailed chemical and physical simulations, faster solutions for certain optimization problems, and new approaches to machine learning.

Cryptography experts are focused on Shor’s algorithm, published in 1994, which can solve the discrete logarithm problem that underlies many public-key systems. Bitcoin currently relies on the practical absence of a quantum device large and stable enough to run a full key-recovery attack.

Recent technical analysis from Google lowered previous qubit estimates for a full Bitcoin-targeting attack to under 500,000 physical qubits and described a staged attack path. The analysis notes part of the computation depends only on fixed elliptic-curve data that is public and common to Bitcoin wallets; a quantum device could precompute that portion and stay on standby. Once a public key appears in the mempool during a transaction or on-chain from a prior spend, the device would only need to complete the second stage. Google estimated that stage could take about nine minutes versus Bitcoin’s roughly 10-minute average block time, creating a short window for an attacker to try to derive a private key and submit a competing transaction. Approximately 6.9 million bitcoins are held in addresses whose public keys have already been exposed, which would be vulnerable to an at-rest quantum attack if capable machines exist.

Private quantum companies and chipmakers are attracting investor interest. Xanadu Quantum Technologies, which builds photonic quantum devices that use light transmitted through fiber, plans to develop a quantum data center by 2030. The company’s founder saw his stake rise in value after the firm’s public debut. Nvidia released open-source artificial intelligence models intended to support research combining AI methods with quantum computing.

Researchers and engineers note many technical challenges remain, including scaling qubit counts, maintaining coherence and implementing error correction. Banks and institutional investors are managing current cost pressures while maintaining selective investments and research programs in quantum computing.