A group of researchers just released a paper showing that quantum computers can generate certified randomness, with significant implications for fields like cryptography.
Published in Nature, the paper highlights random-number generation as a key area where quantum computing outshines classical systems. Quantum mechanics inherently involves randomness, making it a natural fit for this task. Marco Pistoia, a distinguished engineer at JPMorgan Chase, emphasized that certified randomness can be applied in various areas, including cryptography, complex problem-solving, and ensuring fairness and privacy.
The paper also points out a crucial challenge: any device receiving randomness from an external source, like a hardware security module, must verify that the randomness is genuine and recently generated. To address this, the research team from JPMorgan Chase, Quantinuum, Argonne National Laboratory, Oak Ridge National Laboratory, and the University of Texas at Austin used a method called Random Circuit Sampling (RCS). RCS can produce more randomness in its output than in its input, showcasing a capability that classical computers can’t match.
Scott Aaronson from the University of Texas noted that when he proposed his certified randomness protocol in 2018, he didn’t expect to wait so long for a practical demonstration. The team utilized a 56-qubit Quantinuum System Model H2 quantum computer to show it could surpass the computational abilities of today’s most powerful classical supercomputers. By accessing the H2 over the internet, they generated certifiably random bits. They verified this randomness using some of the leading supercomputers, like Frontier and Summit, which are equipped with GPUs ideal for simulating quantum circuits. This setup allowed the researchers to certify 71,313 bits of entropy, a measure of randomness.
Aaronson described this breakthrough as a preliminary step toward using quantum computers for generating certified random bits in cryptographic applications. Pistoia remarked that this accomplishment represents a significant milestone, demonstrating how quantum computing can tackle real-world challenges beyond current classical capabilities.
Rajeeb Hazra, CEO of Quantinuum, noted that this work sets a new standard for quantum security and opens doors for advanced simulations in industries such as finance and manufacturing. Travis Humble from Oak Ridge National Laboratory highlighted that these quantum computing advancements were made possible by the leading-edge computing facilities of the US Department of Energy, pushing the boundaries of what’s possible in computing and shedding light on the convergence of quantum and high-performance computing.