Sandia National Labs scientists invent new benchmarks for measuring performance.

What do a quantum computer and a top sports draft pick have in common? Both have drawn a lot of attention from talent scouts. Quantum computers are experimental machines that can do some tasks faster than supercomputers. Talent scouts are constantly evaluating them for their potential to be a game-changing technology one day.

Scientist scouts now have the first tool to evaluate a potential technology’s ability to run practical tasks. This will reveal its true potential and limitations.

Sandia National Laboratories has developed a new benchmark test that predicts the likelihood of a quantum processor running a program without errors.

Mirror-circuit is a method that scientists use to develop technologies that will lead to the first practical quantum computer. This could accelerate research in medicine, chemistry, and national security.

Scientists have measured performance on obstacle courses that are random operations up to now.

Sandia National Laboratories created a more efficient and accurate method of testing quantum computers. It is pictured here. Credit: Bret Latter, Sandia National Laboratories

New research shows that conventional benchmark tests are underestimating quantum computing errors. This can lead to unrealistic expectations about the power and utility of a quantum computer. According to the paper, mirror circuits provide a more precise testing method.

A mirror circuit is a computer program that reverses the results of a set calculation.

“It’s a common practice in quantum computing to use random, disordered programs to measure performance. Our results show that this is not a good idea,” stated Timothy Proctor, a computer scientist from Sandia’s Quantum Performance Laboratory, who was part of the research.

This new method of testing will allow researchers to evaluate more sophisticated machines. It also saves time. By running the same set of instructions on both a quantum machine and a regular computer, benchmark tests can check for errors. The results should match if there are no errors.

Researchers can wait for quantum computers to complete specific calculations, but they are much more efficient than regular computers.

However, a mirror circuit will always produce the same output as the input or with some intentional modification. Scientists can now instantly check the quantum computer’s results instead of waiting.

The research was supported by Sandia’s Laboratory Directed Research and Development program and the Department of Energy’s Office of Science. Sandia is a crucial Quantum Systems Accelerator (a Department of Energy national quantum research center) member.

A new method exposes flaws in traditional performance ratings.

Proctor and his colleagues discovered that randomized tests underestimate or miss the compound effects of errors. A compounded error can make it worse. It’s like a wide receiver running the wrong route. As the play progresses, they stray farther from their intended destination.

Sandia discovered that functional programs could be imitated, and the final results are often more inconsistent than random tests.

Proctor stated, “our benchmarking experiments revealed the quantum computers’ performance on structured programs is much more variable than previously thought.”

Scientists also have greater insight into using the mirror-circuit method to improve quantum computers.

Proctor stated that by applying our method to quantum computers’ current, we learned much more about the errors these devices suffer. Different types of errors affect other programs in different ways. These effects were not previously observed in many-qubit processors. This is the first method to probe these error effects at a large scale.