The convergence of HPC, AI, and data analytics has driven many infrastructure teams to consider supporting workloads for both compute centric and data centric uses with a single pool of resources. However, some infrastructure teams need clusters focused on traditional HPC workloads for simulation and modeling in a separate, siloed environment.

Customers spend significant time designing, deploying, and tuning HPC clusters given the myriad of variables they must consider including network architecture, storage architecture, memory, file systems, operating systems, configuration settings, servers/CPUs, accelerators, runtime libraries, user-level libraries, workload managers, applications, and benchmarking.

Customers would benefit from a complete tool set to reduce architecture design time, deployment time, and management costs while enabling agility for heterogenous environments.