Microsoft has been an AI-centric company for a while now. Most people know of Microsoft"s push into adding AI in software products like Copilot for Windows. However, Microsoft is also offering its AI services for scientific research. This week, the company announced it is partnering with the Pacific Northwest National Laboratory (PNNL) in Richland, Washington in a project designed to make a new and better lithium battery.
In a blog post, Microsoft says that its Azure Quantum team used its AI and high-performance computing (HPC) resources to help PNNL to help create a new battery material. Microsoft"s AI systems helped to speed up the process of finding 500,000 possible stable materials for this new battery in just a few days.
Microsoft said:
AI may be fast, but it isn’t perfectly accurate. The next set of filters used HPC, which provides high accuracy but uses a lot of computing power. That makes it a good tool for a smaller set of candidate materials. The first HPC verification used density functional theory to calculate the energy of each material relative to all the other states it could be in. Then came molecular dynamics simulations that combined AI and HPC to analyze the movements of atoms and molecules inside each material.
Using HPC, Microsoft and PNNL cut down the possible battery material list to just 150 candidates. It continued to use HPC to further reduce the list by adding in filters like availability, cost, and other factors. In the end, the combination of AI and HPC allowed Microsoft and PNNL to bring down the battery material list to just 18 candidates in the space of only 80 hours.
As a result, PNNL was able to save tons of time to find suitable materials for its battery prototypes. The final result was a material that combines lithium, sodium, and other elements. PNNL says the new material could cut back on the amount of lithium used in a typical battery by up to 70 percent.