The ability to produce random numbers is crucial for a variety of high-performance computing and artificial intelligence research, as well as the encoding of secure data and simulations of real, randomized events. However, classical computers are only able to produce pseudo-random numbers: computers only work by following specific commands, so there’s no way for them to be truly random.
Quantum mechanics is naturally truly random. For that reason, quantum systems are the ideal producers of completely random numbers.
“Classical instruments tend to be cyclic, which means they cycle over the same set of points, so you don’t get a true representation of the system if you’re trying to do computing simulations,” said Carol Scarlett, founder of Axion Technologies. “Quantum mechanics, as far as we know, is the only source of true randomness, and we work to provide that using our devices.”
Axion Technologies is a hardware company that provides quantum random number generators. Their generator devices are made to plug into high-performance computers, producing truly random numbers as the computers require them. Having truly random numbers allows these computers to fully explore all the potential possibilities in their simulations or machine learning algorithms, where the pseudo-random numbers produced by a classical computer might not have. In addition, it has applications in cybersecurity to make key distribution more secure.
Scarlett is an experimental particle physicist who originally worked on searches for dark matter. Her company, “Axion,” is named after an elusive theoretical particle that could potentially explain the cosmological mystery of dark matter, if ever detected; it was while working on an axion detection method that Scarlett invented the quantum device her company uses to generate random numbers.
“I was designing an experiment where you bounce light back and forth in a mirror cavity, and the light would couple and decouple to axionic particles if they passed through,” Scarlett said. “Then I realized I could design a material system to mimic that behavior using photonics. So Axion Technologies grew out of using photonics and real material to mimic what would happen to photons passing through dark matter axions—if they exist, of course.”
The quantum random number generator works by passing photons through birefringent material—birefringence divides the incoming photons into two different polarizations, at completely random rates. Axion’s device can create a large stream of randomized bits, which is what most computing applications require from a random number generator.
“True randomness is a huge benefit to cybersecurity systems, and a great asset for modeling how complex systems in nature really work,” Scarlett said. “It assures you that your results are going to be a good representation of what’s truly out there.”
Axion is part of Argonne National Laboratory’s Chain Reaction Innovations program, embedded there to make a chip-scale device that can plug into other kinds of computer-like systems, such as smartphones and other smart devices.