New QUOPS Benchmark Quantifies Quantum Computing Usefulness — SkimNews
Get the Health newsletter
Daily health & science — research, biotech, public health, the studies worth knowing. Free.
- Timothy Proctor at Sandia National Laboratories and colleagues developed the QUOPS (quantum universal operation performance system) score, which combines qubit count and computational operations to gauge whether a machine can run a given algorithm.
- QUOPS scores for leading quantum computers from Google, IBM, and Quantinuum all fell below 2,000 — roughly 100,000 times lower than what running useful FeMoco (nitrogen-fixing) or cryptography-breaking computations would require.
- Quantinuum's Helios-1 configured for fault-tolerant operation — which catches and corrects its own errors — earned a QUOPS score of only about 40, underscoring how far fault-tolerant quantum computing still has to go.
- Sam Stanwyck at Nvidia, which collaborated on the project, called QUOPS a "reality check" and said researchers could use it to see how small hardware tweaks shift scores, accelerating progress toward useful machines.
- Wolfgang Mauerer at the Technical University of Applied Sciences Regensburg said QUOPS could evaluate quantum computers using very different qubit types, but warned the metric can't capture everything that makes a quantum machine powerful.
- Proctor identified the field's inflection point as the moment logical qubits — used for fault-tolerant computing — begin scoring higher on QUOPS than ordinary qubits do today.
Why it matters: Today's top quantum computers score below 2,000 on the new benchmark, while real-world problems like cryptography-breaking and FeMoco simulation demand roughly 100,000 times more — a concrete yardstick researchers and hardware makers (Google, IBM, Quantinuum, Nvidia collaborators) can now use to track whether each new generation is actually closing that gap.
Ask SkimNews
