Waterloo researchers in the Department of Physics and Astronomy and the Institute for Quantum Computing are on the fast-track to developing high-performance software to assist with cybersecurity risk in quantum computing.
“Much of the world is going to need programmers and problem solvers,” said Olivia Di Matteo, a second-year Waterloo Physics doctoral student at the University of Waterloo. Di Matteo spent the past two years of her master’s degree working on an advanced computing project with U of Waterloo professor and principal investigator, Michele Mosca.
The project involved developing software for quantum circuit synthesis, a process that “translates” arbitrary quantum operations into a set of instructions that can be run on a quantum computer. The project supports a start-up company called evolutionQ, which provides cybersecurity risk assessments.
The research is timely. With online activity on the rise, Canada will indeed require more adept problem solvers to develop secure protocols and encryption software. Canadians are sending more emails than ever, both personally and professionally, and there has been a steady increase in other online activities such as online banking, online shopping, e-health, and cloud computing, all of which puts sensitive data at risk.
Di Matteo, a Thunder Bay native and recipient of an NSERC Canada Graduate Scholarship for her PhD studies, became interested in parallel computing through undergraduate summer internships. Although she’s worked on advanced computing platforms previously, her work with Prof. Mosca was her first experience with a project of such scale.
The team quickly discovered they needed access to SOSCIP’s advanced computing platforms to deliver the results they were looking for, namely the Blue Gene/Q (BGQ), Canada’s fastest supercomputer.
“There weren’t enough processors to solve the problems that were interesting to us. The BGQ allowed us to get really good results. The software is largely parallel, so we needed a lot of time and a lot of processors. Because we were able to get so many processors working together, it was really beneficial – our run time changed from a week to a couple of minutes.”
To put that in perspective, the biggest program the team ran took 4000 processors and 21 hours, a project which would have taken weeks or longer on standard systems.
Di Matteo also credits the expert knowledge of the SOSCIP team with furthering the research.
“I had a little training in high-performance computing but not in parallel computing and we needed to learn how to make all of the processors talk to each other effectively. The technicians were really helpful in understanding how to best take advantage of the technology. If I got stuck and didn’t know if my method was efficient, I had someone to bounce ideas off and get suggestions.”
Today, the software is a complete and available for researchers to download. The research informs evolutionQ, a start-up co-founded by Waterloo professors Michele Mosca and Dr. Norbert Lütkenhaus. It also provided valuable training skills for Di Matteo and others students in high performance computing, a skill which is going to be in high demand as quantum computers become the norm.
“The software might be useful in the future for people working on quantum computers and building implementations,” explained Di Matteo.
For Mosca, the time is now to develop the cryptographic infrastructure needed to protect our data.
“The work [of our research project] directly impacts cybersecurity, since the currently deployed cryptographic infrastructure, which underpins cybersecurity, will be compromised by a large scale quantum computer. It is imperative that this quantum threat is mitigated before it becomes a reality,” explained Mosca.
“The suite of cryptographic tools we rely on to provide cybersecurity must evolve to a suite of quantum-safe cryptographic tools.”
Researcher: Prof. Michele Mosca, University of Waterloo and Olivia Di Matteo, PhD student
Project title: A parallel algorithm for quantum circuit synthesis
Supported by: SOSCIP, Federal Economic Development Agency for Southern Ontario, IBM Canada Ltd., University of Waterloo