New Benchmark Finds Quantum Computers Still Far From Scientific Utility
A New Benchmark Suggests Today’s Leading Quantum Computers Need Roughly 100,000 Times More Computational Capability to Reach Major Scientific Applications.
Researchers have introduced a new benchmark designed to provide a more meaningful way of comparing the computational capabilities of quantum computers. The system, called the Quantum Universal Operation Performance System, or QUOPS, measures both the size and speed of quantum circuits that processors can successfully execute. The researchers applied the benchmark to leading systems from Quantinuum, Google and IBM.
The study found that current quantum processors remain substantially below the computational capability required for recognised scientific challenge problems. According to the researchers’ analysis, quantum systems would need to increase their effective computational capability by approximately five orders of magnitude—around 100,000 times—to reach the level required for these applications.
The significance of the benchmark is that it looks beyond simply counting qubits. A machine can have a large number of qubits but still struggle with useful calculations because of errors, limited circuit depth and other hardware constraints. QUOPS instead attempts to capture how large and complex a computation a system can actually perform successfully, while also accounting for execution speed.
Researchers also tested the benchmark on a fault-tolerant logical-qubit processor implemented using Quantinuum’s Helios system. The results provide a framework for tracking how future generations of quantum computers could progress toward practical scientific applications, rather than relying solely on hardware specifications such as raw physical-qubit counts.
The findings do not suggest that quantum computing has failed; instead, they provide a clearer picture of the distance still to travel. Quantum researchers continue to improve hardware, error correction, algorithms and system architecture, with potential applications ranging from materials science to chemistry and drug discovery. The new benchmark could become an important tool for determining whether those advances are translating into genuinely useful computational power.
