Edge Computing in AIMD
The Edge Computing Module facilitates the processing of large and high velocity data streams from the different subsystems, analyzes them through machine learning (ML) algorithms in real-time, and provides fast feedback to the different subsystems. This is achieved by having all subsystems connect to a local node (head director) via a 100 or 200gbps. The head director is connected to a 1 PB NVMe local storage for data for real time analysis and decisions. The head director has a field-programmable gate array (FPGA) and four graphics processing units (GPU) that accelerate the local analysis and learning of the data. The head director has another 100gbps IB card for networking to MARCC/ARCH for long-term storage and more intense analysis and learning.
The Advanced Research Computing at Hopkins (ARCH) -formerly known as MARCC- is a shared computing facility at Johns Hopkins University that enables research, discovery, and learning, relying on the use and development of advanced computing. ARCH administers state of the art high performance computing resources, manages highly reliable data storage, and provides outstanding collaborative scientific support to empower computational research, scholarship, and innovation.
Visit the ARCH Website here: https://www.arch.jhu.edu/