Researchers just made a battery breakthrough in record time using AI and cloud computing. Microsoft teamed up with scientists at Pacific Northwest National Lab to discover promising new battery materials in a matter of days rather than years.
They leveraged advanced AI running on Microsoft’s Azure cloud platform to analyze half a million stable candidate materials in a few days. This process traditionally takes painstaking trial-and-error testing by humans.
AI algorithms swiftly generated and evaluated millions of molecular combinations. Then high-performance computing refined the list for optimal properties like conductivity and stability.
The scientists say AI brought speed while HPC delivered precision. Together they narrowed the options down to just 23 candidates worth pursuing in the lab – a task that took only 80 hours thanks to cloud computing power.
One material they synthesized replaces up to 70% of lithium with sodium and other elements. It acts as a solid electrolyte to help ions flow efficiently between electrodes.
While still early days requiring optimization, it highlights the unprecedented pace of discovery possible with AI. In the past, developing radically new battery chemistries took years or decades.
The goal is to eventually simulate the full process from molecular design to battery performance digitally. This “digital twin” concept would minimize physical experiments.
The lead scientist called it a new era accelerating solutions to global energy problems. By combining AI’s raw speed with HPC’s accuracy, they can rapidly filter huge materials datasets for needle-in-a-haystack discoveries.
While most projects still require painstaking lab validation, digital techniques are turbocharging the pace of innovation. This latest battery breakthrough was facilitated by cloud-based AI, allowing computing power and algorithms to do the heavy lifting.
