(INSIDE HPC) – At last week’s ISC 2025 conference in Hamburg, we spoke with Prof. Volker Lindenstruth, Director of the Center for Scientific Computing at Goethe University, Frankfurt. Researchers at the university are leveraging one of Europe’s largest AMD GPU clusters paired with VDURA’s recently announced V5000 all-Flash NVMe storage appliance to accelerate AI-driven physics and digital twin models.

The V5000, built for AI factories, is designed to meet the requirement of high-demand AI pipelines as generative AI models move into production. Built on the VDURA V11 data platform, the system delivers GPU-powered throughput “while ensuring the durability and availability of data for 24x7x365 operating conditions, setting a new benchmark for AI infrastructure scalability and reliability,” VDURA said.

In this interview, Volker discusses the performance of the VDURA- AMD combined solution, including:

• The demands of large-scale AI and physics workloads
• Why the university selected VDURA as their storage solution
• The importance of multi-level erasure coding and data resilience at petabyte scale
• Real-world performance metrics: 8.4 million IOPS and 670K metadata ops
• How VDURA enables cost-effective, university-scale storage modernization