Case Study: University of Florida achieves world-class AI supercomputing with DDN and NVIDIA DGX SuperPOD

University of Florida reduces processing latency up to 9X with DDN

The University of Florida faced the challenge of upgrading its HiPerGator supercomputer to support its ambitious goal of providing an AI curriculum for every student. Their existing infrastructure struggled with rapidly expanding research data and the conflicting computational demands of multiple university departments, which risked slowing research and impacting the user experience. To overcome this, they partnered with vendor DataDirect Networks.

DataDirect Networks, alongside NVIDIA, implemented a solution using the DDN A³I AI400X all-flash storage system as part of an NVIDIA DGX SuperPOD. This significantly boosted the university's processing capacity to 0.7 Exaflops and reduced processing latency for end users by up to nine times. The solution provided the scalable, high-performance infrastructure needed, propelling HiPerGator 3.0 to become the world's fastest AI supercomputer in academia.

University of Florida