Real World Application, Performance and Optimization on the Marconi Supercomputer.
Carlo Cavazzoni (CINECA‘s R&D Director for HPC optimization strategies) made a live Intel® Omni-Path Architecture Webinar on July 12, 2018.
In the Webinar he explained how Cineca successfully ported and optimized Quantum ESPRESSO (QE) on the Marconi supercomputer, deploying QE on one of Europe’s fastest supercomputers.
Marconi is based on Lenovo ThinkSystem servers and storage systems powered by Intel® Xeon® Scalable and Intel® Xeon Phi™ processors interconnected with Intel® Omni-Path Architecture 100G fabric.
The webinar users learned about:
- Cineca’s optimization strategies
- Benchmarks on QE performance on Marconi
- Insights on how Cineca and collaborators are facilitating material science transition to exascale
- Details on a live QE and field test at an upcoming MaX Hackathon