EMbeDS INFRASTRUCTURE: SECOND PHASE OF HIGH-PERFORMANCE COMPUTING ACQUISITIONS COMPLETED

After 9 months of usage of our new HPC infrastructure, we analyzed the additional needs matured by EMbeDS researchers and planned enhancements comprising:

• a stand-alone Server for public services (separate from the HPC cluster to ensure data safety)
• network equipment for fault tolerance and increased connectivity
• increased RAM memory on existing nodes
• a High Computational Density node

These further acquisitions consolidated our infrastructure, making it fault-tolerant and able to offer new functionalities in support of different types of research projects (utilities, memory-intensive, CPU-intensive).

In more detail, we added 768 GB RAM to upgrade the performance of the two existing computational nodes (DELL PowerEdge R740), and included two Dell EMC Connectrix DS-6610B Switches (one for redundancy) for connectivity to the Storage Area Network of the Sant’Anna School. We then enhanced our HPC with a new High Computational Density node DELL PowerEdge R840 (4x Intel® Xeon® Gold 6252, 768 GB RAM) integrated with the VMware vSphere 6.7U3 virtualization system, which enables support for memory-intensive applications. Finally, we integrated a new Ethernet Dell EMC Switch S4128F-ON for network redundancy and a Stand-alone Dell PowerEdge R6525 Server (2xAMD EPYC 7452, 64C, 256 GB RAM) to support research projects of the Department of Excellence that provide public services.