Case Study: Lawrence Berkeley National Laboratory accelerates energy policy research with Coiled

How Lawrence Berkeley National Laboratory processes 55 years of weather patterns

Lawrence Berkeley National Laboratory, a research lab for the Department of Energy, was faced with the immense computational challenge of processing petabytes of weather and land-use data for critical energy policy research and grid planning. Their local machines and traditional HPC clusters were unable to handle the workload efficiently, while direct cloud computing was cost-prohibitive and complex to manage. This impeded their ability to provide timely analysis to policymakers on America's clean energy transition.

By implementing Coiled, the lab gained access to a Python-native cloud computing platform that integrated seamlessly with their Dask and Xarray workflows. Coiled enabled them to easily scale their computations to hundreds of cores on-demand, drastically reducing analysis time and eliminating the frustrations of infrastructure management and cost overruns. This solution empowered researchers to rapidly deliver crucial insights on policy impacts and develop interactive tools for identifying renewable energy potential, directly supporting the nation's shift to clean energy.

Lawrence Berkeley National Laboratory

Umed Paliwal

Researcher