Intelligent data. Unmatched results. Activate your AI Advantage with NetApp- Yoshiki Fujiwara | Hitachi
How can researchers use cloud compute when hundreds of terabytes of research data remain on premises? This session explores a hybrid HPC proof of concept by RIKEN’s Brain Connectomics Imaging Research Team, NetApp, Astec and Fusic. Learn how Amazon FSx for NetApp ONTAP and NetApp FlexCache make requested data available near AWS compute without a full dataset pre-copy, while AWS ParallelCluster and Slurm orchestrate CPU and GPU resources. We will distinguish first-read and cached-read behavior, share practical lessons on network validation and workload-based sizing, and explain why compute elasticity and storage cost management require separate decisions. The session separates demonstrated small-scale results from future scale-out plans and offers a practical starting point for your own hybrid HPC evaluation.