If quantum computers solve unsolvable problems, how can their answers be verified?

New research will let scientists to use classical laptops to efficiently verify whether quantum computer outputs match expected results.

While quantum computers can solve hyper-complex problems in record time, how can researchers ever verify efficiently, if the solutions are correct?

Traditional computers would take thousands or even millions of years to analyze the same problems that quantum machines tackle in minutes, so direct comparison is not feasible.

On 17 September 2025, scientists at Swinburne University of Technology, have announced a way to use Gaussian Boson Samplers (GBS) — devices that process photons to calculate complex probability distributions — to check the accuracy of certain quantum computers, addressing one of the field’s biggest puzzles: how to verify solutions to problems that classical computers cannot possibly solve in any reasonable time frame. Potentially, this validation process can ensure the reliability and trustworthiness of quantum applications ranging from drug discovery to cryptography.

The Swinburne group’s validation techniques uses “phase-space simulations” to quickly check, on a standard laptop, whether a quantum computer’s results match the expected distributions.

So far, their tests, which are exponentially more efficient than classical checks, have identified substantial discrepancies in outputs from a recent GBS experiment. The quantum device’s findings did not line up with theoretical predictions, revealing extra noise and other errors affecting the result.

This raises an even deeper question for quantum computing: When machines claim to solve problems that are otherwise unsolvable, how can we be confident that quantum answers are workable? As quantum computers move closer to commercial and industrial deployment these new methods are crucial to ensure “fault-tolerant” quantum machines capable of reliably harnessing quantum effects by 2029. Breakthroughs in quantum validation are essential to ensure these machines produce results that are not just fast but correct.

The Swinburne research concludes that scalable, quickly deployable validation tools must be at the heart of future quantum computing development, providing feedback on experimental performance and flagging sources of error.

As quantum computing becomes part of the real world — potentially transforming science, technology, and industries — the ability to trust and check quantum answers will determine how broadly this new technology is adopted.