A new study proposes a framework to evaluate when quantum computers can outperform classical ones in combinatorial optimization problems, a crucial area for quantum utility. Researchers have identified three key milestones that must be reached to demonstrate practical quantum advantage, and have developed a set of benchmark problems to test the performance of quantum algorithms. This work is fundamental to guide the development of quantum computing towards real-world applications, beyond theoretical advantage demonstrations.

The first milestone refers to the ability of quantum algorithms to find high-quality solutions for optimization problems. The second milestone assesses whether these algorithms can solve larger problems than what is feasible for classical methods. Finally, the third milestone focuses on efficiency, i.e., whether quantum algorithms can achieve these high-quality solutions for large problems faster or with fewer resources than their classical counterparts. These milestones provide a clear path for research and development in quantum computing, allowing for an objective evaluation of progress.

To facilitate this evaluation, the team has created a set of benchmark problems spanning diverse structures and complexities, from constraint satisfaction problems to maximum cut problems. These problems are designed to be scalable and to allow for fair comparisons between different quantum architectures and classical algorithms. The proposed methodology enables researchers to quantify "quantum utility" in terms of solution quality, scalability, and computational efficiency, offering a standardized metric for the field.

This framework is crucial for the transition of quantum computing from fundamental research to practical applications. By clearly defining what constitutes a useful quantum advantage, the study helps focus development efforts on building systems that can address real-world problems more effectively. These benchmarks are expected to drive innovation and accelerate the arrival of quantum computers capable of solving complex challenges in fields such as logistics, materials chemistry, and drug discovery.