The University at Buffalo has chosen Panasas ActiveStor Ultra HPC data storage for the university’s Center for Computational Research (CCR), a supercomputing center serving the 64 campuses of the State Universities of New York (SUNY) network and affiliated partners, including industry.
In its announced, the university (UB) said it moved to Panasas because it was challenged by higher quantity and complexity of mixed STEM workloads — “in addition to support and stability issues” with its existing storage infrastructure. Panasas has provided 1.5 petabytes of storage using ActiveStor Ultra, a turnkey HPC storage appliance that runs the PanFS parallel file system, designed “to accelerate performance at every stage of the computational research process,” according to Panasas.
“Supporting the high-performance computing needs of a group as diverse as faculty, staff and students – as well as local business and industrial partners – is a massive task, and the University at Buffalo and its Center for Computational Research are meeting that challenge every day,” said Jim Donovan, chief marketing officer at Panasas. “At Panasas, we believe storage technology should support innovation with a seamless and uninterrupted experience that optimizes productivity and time-to-quality outcome. We are proud to support UB and its mission of academic excellence.”
Panasas storage features Dynamic Data Acceleration, designed to automatically adapt to the changing small file and mixed workloads in HPC and AI. “Dynamic Data Acceleration eliminates the complexity and manual intervention of tiered HPC storage systems by maximizing the efficiency of all storage media in a seamless, total-performance system that automatically adapts to changing file sizes and workloads,” Panasas said.
In this integrated system, NVMe SSDs store metadata, low-latency SSDs store small files, and large files are stored on low-cost, high-bandwidth HDDs, according to Panasas. “By dynamically managing the movement of files between SSD and HDD, and maximizing the full potential of NVMe, PanFS delivers the highest possible performance at the lowest cost for HPC and AI workflows,” the company said.
