An international team of researchers has achieved a significant breakthrough by successfully generating certified random numbers using a quantum computer, marking a pivotal advancement with implications for cybersecurity and various industries. This achievement addresses the inherent limitations of traditional random number generators, which, despite appearing random, rely on deterministic processes.
The research, published in Nature on March 26, 2025, involved scientists from JPMorgan Chase, Quantinuum, Argonne National Laboratory, Oak Ridge National Laboratory, and the University of Texas at Austin. They utilized Quantinuum's System Model H2, a 56-qubit trapped-ion quantum computer, to perform a randomness-expansion task that surpasses the capabilities of classical supercomputers.
This quantum system's ability to generate entropy beyond classical reach was validated using over 1.1 exaflops of classical computing power. The team employed a technique called Random Circuit Sampling (RCS) to expand randomness, outputting more randomness than it takes as input. This is particularly crucial for creating unpredictable cryptographic keys, thereby bolstering cybersecurity defenses against increasingly sophisticated cyberattacks, including potential threats from future quantum computers.
The demonstration, conducted via the internet, highlights the accessibility and potential for widespread adoption of certified quantum randomness. Experts suggest that this breakthrough not only demonstrates the unmatched performance of trapped-ion technology but also sets a new standard for robust quantum security and enables advanced simulations across diverse sectors like finance and manufacturing. This marks a significant stride in transitioning quantum computing from theoretical possibilities to practical, real-world applications.