Microsoft recently announced its collaboration with the U.S. Department of Energy’s Pacific Northwest National Laboratory (PNNL) in utilizing its Azure Quantum Elements service. This collaboration successfully narrowed down millions of potential battery materials to just a few, with one now progressing to the prototype phase.
It’s important to clarify that despite the allure of the term “quantum” in “Azure Quantum Elements,” this project did not involve a quantum computer. Launched last summer, Azure Quantum Elements integrates AI with traditional high-performance computing (HPC) techniques. This service serves as a platform for scientific computing, with the eventual goal of providing access to Microsoft’s future quantum supercomputer. So, while quantum technology wasn’t directly used in this project, the overarching goal is to eventually converge these various technologies.
Krysta Svore, head of Microsoft Quantum, explained that the objective was to maximize the potential of Azure Quantum Elements (AQE), particularly its AI accelerator, in advancing material discovery. Researchers at PNNL analyzed 32 million inorganic materials using AQE, identifying 18 promising candidates for their battery project. The process began with AI models in AQE reducing the selection to around 500,000 possibilities. Following this, HPC techniques helped pinpoint the final 18 candidates. Typically, such a process would span years, including prototype development, but with AQE, it was accomplished in just 18 months.
Tony Peurrung, PNNL’s Deputy Director for Science and Technology, emphasized the importance of merging AI, cloud, and high-performance computing with human expertise to expedite meaningful scientific outcomes. This collaboration with Microsoft aims to make AI more accessible to scientists, potentially uncovering unconventional but worthwhile materials and approaches. This project marks the beginning of what is expected to be a significant journey in accelerating scientific discovery.
Quantum computing is anticipated to greatly enhance chemistry and material science problem-solving. However, the quantum computing field is still progressing and is a few years away from realizing a truly useful quantum computer. We are presently in the noisy intermediate-scale quantum (NISQ) era. Svore is optimistic about Microsoft’s plan to develop a quantum supercomputer using Majorana-based qubits within the next decade.
At present, even with the tangible scientific work involved, one could perceive this as somewhat of a PR move, given the current distance from integrating quantum computing into such processes.