Updated
Updated · New Scientist · Sep 15
QUOPS Benchmark Shows Top Quantum Computers Trail Useful Tasks by 100,000-Fold
Updated
Updated · New Scientist · Sep 15

QUOPS Benchmark Shows Top Quantum Computers Trail Useful Tasks by 100,000-Fold

3 articles · Updated · New Scientist · Sep 15

Summary

  • Google, IBM and Quantinuum machines all scored below 2,000 on the new QUOPS benchmark, far short of the level needed for major tasks such as FeMoco simulation or breaking current cryptography.
  • Timothy Proctor's team designed QUOPS to compare a machine's usable qubits and executable operations against an algorithm's required score; a separate QUOPS rate estimates how long the computation would take.
  • Quantinuum's fault-tolerant Helios-1 configuration scored about 40, highlighting how error-correcting systems remain early even as researchers see fault tolerance as quantum computing's only viable path to scale.
  • The benchmark is intended as a reality check and design tool, letting researchers test how hardware tweaks change capability, though outside experts say no single score will fully capture quantum computing power.

Insights

Can QUOPS finally expose the true capabilities of quantum hardware, or will the speed-versus-capability tradeoff keep the ultimate winner a mystery?
Will the hidden limits of ion shuttling eventually bottleneck Quantinuum’s massive leap in fault-tolerant quantum memory?
With Helix merging different error-correcting codes into one architecture, are we secretly witnessing the end of the raw qubit-count race?