Exascale workflows
MaX designs and implements exascale-oriented workflows aimed at optimizing computational resources for complex simulations. These workflows drive advances in materials science as they can harness the full potential of exascale computing.
Workflows enable complex, multi-step simulations, allowing the efficient coordination of large-scale tasks while managing the data flow. In addition, workflows help maintain a unified data store. This minimizes the need for extensive data transfers and overcomes traditional resource limitations.
At MaX, we build and connect tools that handle workflow design, orchestration, scheduling, data sharing, and code compatibility. These tools work together to run reliably on exascale systems. The result is scalable, automated, and reproducible simulations that maximize HPC at exascale.
MaX workflow tools
MaX builds workflows for exascale computing to enable codes to run automatically and fully exploit HPC power. Its core tools are AiiDA, which manages workflows and tracks provenance, and HyperQueue, which ensures efficient task distribution. Additional tools, including ZeroMQ for data exchange and various format converters, enhance workflow interoperability.
Exascale Workflows and Extreme Data for Materials
Workflows address key challenges in materials science, from thermal conductivity to nanoscale magnetism and battery design. They combine different codes and methods to perform analyses that single simulations cannot achieve. This approach drives progress in energy, storage, and quantum technologies. Learn more in our infographic.