At a Glance
- Scientists at Los Alamos National Laboratory identified a new computational task that is exceptionally difficult for classical computers but can be solved efficiently using a quantum computer.
- This problem involves simulating an extremely complex optical network of mirrors and light particles, known as a Gaussian bosonic circuit, on a vast and unmanageable scale.
- Researchers formally proved the problem belongs to a class called BQP-complete, confirming its intractable nature for classical computers while remaining perfectly solvable for quantum ones.
- To illustrate the power of their framework, the team numerically simulated a massive interferometer system containing approximately 8 billion light modes, a scale far exceeding classical computational limits.
- This work significantly expands the exclusive list of problems demonstrating quantum advantage and deepens our understanding of where the true power of quantum computation lies.
A central challenge in building a useful quantum computer is finding problems where the new technology will have a clear advantage over today’s machines—in a new study published in Physical Review Letters, Los Alamos National Laboratory scientists added a significant new task to that exclusive list. The team developed a method for a quantum computer to efficiently simulate a vast and complex optical system, a feat considered intractable for even the most powerful classical supercomputers.
The problem involves simulating what are known as Gaussian bosonic circuits. These can be imagined as intricate networks of semi-transparent mirrors and other optical components that guide and alter the paths of countless light particles or photons. The sheer scale of the system makes it a computational nightmare for a classical device. “Just writing down a complete description of this system on a classical computer would require an enormous amount of memory and processing capability,” said Diego García-Martín, a co-author with the lab’s Information Sciences group, in a Phys.org press release. The Los Alamos framework bypasses this by encoding the properties of the light system onto qubits, the fundamental building blocks of a quantum computer, and then using a quantum algorithm to simulate how the light evolves.
More than simply demonstrating a simulation, the researchers rigorously proved its value. They showed that the task belongs to a class of problems known as bounded-error quantum polynomial time complete, or BQP-complete. This is a formal way of saying the problem is provably difficult for classical computers but efficiently solvable by a quantum computer. “One of the central questions that [face] quantum computing is what classes of problems they can most efficiently solve but classical computers cannot,” said the team’s lead scientist, Marco Cerezo. “At the moment, this is the Holy Grail of quantum computing, because you can count on two hands such problems. In this paper, we’ve just added another.”
By establishing this specific simulation as a BQP-complete problem, the research provides a concrete, physically-motivated example of quantum advantage. To illustrate the power of their method, the team performed numerical simulations of an interferometer containing approximately 8 billion light modes, a scale far beyond classical reach. This work expands the set of known applications for quantum computers and deepens our fundamental understanding of their true power, pushing the entire field closer to realizing its transformative potential.
References
- Barthe, A., Cerezo, M., Sornborger, A. T., Larocca, M., & García-Martín, D. (2025). Gate-based quantum simulation of gaussian bosonic circuits on exponentially many modes. Physical Review Letters, 134(7), 070604. https://doi.org/10.1103/PhysRevLett.134.070604
- Los Alamos National Laboratory. (2025, June 12). A new problem that only quantum computing can solve. Phys.Org; Los Alamos National Laboratory. https://phys.org/news/2025-06-problem-quantum.html
