The Center for Data Science maintains a close partnership with Research Computing & Data (RCD) at UMass Amherst. RCD develops and manages scalable, campus-wide research computing solutions to support the high-performance computing and data management needs of researchers.
Unity
Spearheaded by UMass Amherst RCD and located at the MGHPCC, the Unity Research Computing Platform is a collaborative research computing environment built for discovery.
Whether your group needs to write Jupyter Notebooks in a collaborative location, run thousands of ensemble experiments, or scale massively parallel workloads across hundreds of cores, Unity handles the infrastructure so you can focus on the research.
GPU RESOURCES
The Center for Data Science supports priority access to a cutting edge “superpod” of over 100 NVIDIA A100 GPUs designed for training the largest and most powerful AI models.
This specialized resource is configured with NVLink as well as a high bandwidth GPU-direct network enabling ultra-fast training of models across multiple GPUs and nodes in the entire pod.
Featured: Hojae Son
Hojae Son is a joint research facilitator for Research Computing & Data and Center for Data Science. His work focuses on facilitating workflow execution on Unity, where he supports machine learning researchers by setting up multi-node AI training environments.
With expertise in multi-node GPU training, Hojae is integral to helping research groups leverage UMass’ highest-powered GPU resource: the A100 superpod.
