IBM announced that its IBM Fusion platform has delivered one of the first implementations of the NVIDIA AI Data Platform reference design. This deployment features NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs, NVIDIA Networking, and NVIDIA AI Enterprise software.
UT Southwestern Medical Center is the first client to use this new reference design with IBM Fusion. The medical center aims to enhance large-scale training and inferencing capabilities, which allows for handling larger models, more complex datasets, and providing faster semantic query responses.
The NVIDIA AI Data Platform is a customizable infrastructure solution designed to meet the growing needs of artificial intelligence workloads. It integrates enterprise storage with accelerated computing to support AI agents in delivering near-real-time business insights. By combining NVIDIA’s infrastructure with IBM Fusion’s content-aware data services, clients can process, index, and vectorize unstructured and semi-structured data continuously. This prepares data for AI applications and helps organizations derive valuable insights from their information.
IBM Fusion’s integration with NVIDIA NeMo Retriever Microservices enables clients to use advanced data indexing for AI agents, offering a comprehensive stack for developing inferencing capabilities quickly.
UT Southwestern is among the first organizations globally to adopt this approach using IBM Fusion. The platform automatically prepares the institution’s data for AI purposes, supporting larger model training and enabling rapid and accurate responses to complex queries.
IBM Fusion supports several research-focused use cases at UT Southwestern:
– In drug discovery acceleration, researchers use the open-source NVIDIA BioNeMo framework integrated with IBM Fusion to build deep learning models aimed at identifying new biomolecules.
– For patient digital avatars, internal data and medical literature are securely ingested into IBM Fusion. This powers an environment where AI-driven avatars simulate clinical scenarios for student training.
– The Helixa AI Assistant uses research data processed by IBM Fusion to provide contextual support for BioHPC researchers.
“IBM Fusion is engineered to help organizations conduct deep research by unifying applications and access to data across hybrid clouds to help simplify discovery and data activation that is AI-ready,” said Sam Werner, GM of IBM Storage. “Our integrations with NVIDIA is built to help clients unlock hidden insights from data to supercharge AI and analytics initiatives.”
“AI is opening transformative opportunities in healthcare research—from decoding genomics to interpreting complex medical imaging and accelerating translational breakthroughs,” said Anne Hecht, Senior Director, Enterprise Platforms at NVIDIA. “The context-aware IBM Fusion platform, built with the NVIDIA AI Data Platform reference design, is making diverse, multimodal datasets available to researchers and healthcare providers using AI to accelerate discovery.”
