To support growing demand for accelerated computing for research across artificial intelligence (AI), semiconductor technology, quantum science, and beyond, Purdue University expanded its advanced computing resources through two key initiatives: Gautschi-AI and an expansion of the Anvil supercomputer to support the National AI Research Resource (NAIRR). These major technology refreshes, funded by grants from the Lilly Endowment and the National Science Foundation (NSF), are essential components of the Purdue Computes initiative, which aims to elevate the university’s leadership in AI, quantum, semiconductors, computing, and interdisciplinary research.
With Gautschi and Anvil, Purdue is not only enabling faculty and students to tackle larger and more complex research problems but also positioning itself as a national hub for innovation in AI and high-performance computing (HPC).
“Gautschi is a foundational tool to help us reach top-10 status for our Department of Computer Science and build our new Institute for Physical Artificial Intelligence,” said Preston Smith, executive director of the Rosen Center for Advanced Computing at Purdue University. “High-end computer science and AI is a competitive space. Gautschi sets Purdue apart as a destination for both faculty and students.”
Addressing Demand for Scalable, High-Performance Computing
Purdue worked with AI architects from Dell Technologies and NVIDIA to design the Gautschi-AI cluster specifically for the university research community. Architected with the NVIDIA SuperPOD reference design, a blueprint for large-scale, high-performance AI infrastructure, Gautschi features 20 Dell PowerEdge XE9680 nodes, each with eight NVIDIA SXM H100 GPUs. With scalability up to 32 nodes (256 GPUs), an all-flash, 7 petabyte (PB) parallel file system and ultra-fast 400 Gbps NDR InfiniBand interconnects powered by NVIDIA Mellanox, delivering 10.7 petaflops of peak performance, Gautschi offers the power and flexibility needed to train large AI models and analyze massive datasets across disciplines.
Complementing Gautschi, the Anvil NAIRR expansion expands Purdue’s NSF-funded Anvil system with 21 Dell XE9640 nodes, each with four NVIDIA SXM H100 GPUs and 1PB of flash-based object storage. Anvil’s new resources support broader national AI research through the NSF’s ACCESS program and the NAIRR pilot.
“Anvil is part of a national effort to democratize access to AI computing,” Smith noted. “It ensures that researchers across the country can leverage high-end infrastructure to support their science.”
Accelerating Discovery Across Disciplines
With centralized access to powerful AI and HPC infrastructure, more than 260 researchers across 66 Purdue departments, from agriculture and engineering to humanities and business, are using these systems to fuel innovation and solve big problems. Applications include:
- Materials science: Simulating aircraft materials under hypersonic pressures using AI instead of time and computationally intensive physics-based models
- Forestry: Counting the size of tree populations globally using AI-powered satellite image analysis, producing a first-of-its-kind total global number of tree species estimate (73,300)
- Autonomous systems: Training models for self-driving vehicles in a variety of applications
These use cases highlight Purdue’s growing focus on physical AI – a concept that bridges digital and real-world applications of AI. “We’re using AI to solve tangible problems – whether it’s feeding the world through precision agriculture or improving climate resilience,” Smith said.
Advancing AI With Efficiency
Rather than building isolated infrastructure for each research initiative, Purdue’s strategy has long embraced a shared “community cluster” model, where faculty buy into large-scale centralized resources. This approach significantly reduces duplication, lowers costs, and allows graduate students to focus on research instead of IT work.
“By centralizing infrastructure and scaling it for everyone, we can deliver world-class computing to every corner of our university – and do it efficiently,” Smith said. “Our faculty and students can focus on science rather than infrastructure because we’ve eliminated the need to procure GPUs or wait in long queues on the public cloud.”
This central investment in computing infrastructure currently supports researchers in all seven academic colleges at Purdue’s West Lafayette campus, as well as all three regional campuses. This investment has been demonstrated to translate directly to improved institutional outcomes, including new grant awards, Ph.D. degrees awarded, and high-impact publications.
Collaborating to Advance AI Capabilities
“Our long-term relationship with Dell and NVIDIA has been instrumental,” Smith said. “They understand our researchers’ needs and help us design solutions that match our facilities and goals. Their professional services have been a real differentiator.”
“Purdue challenges us to deliver state-of-the-art systems,” said Rob Holtz, a Dell Technologies solutions executive who works with Purdue. “They’re pushing the boundaries of research and teaching us along the way – this partnership is as much about learning as it is about delivering.”
Supporting Scientific Discovery
Both Gautschi and Anvil are listed on the TOP500 ranking of the world’s most powerful computer systems, giving Purdue an edge in attracting talent, funding, and national recognition. But beyond the numbers, these initiatives are about enabling researchers to ask – and answer – bigger questions.
“Without systems like these, researchers can’t do the science that they are expected to do today,” Smith said. “The investment in centralized, scalable AI infrastructure is an investment in future discoveries.”
Learn more:
Dell Technologies and NVIDIA can help you leverage AI to drive innovation and achieve your business goals. The Dell AI Factory with NVIDIA is the industry’s first and only end-to-end enterprise AI solution, designed to speed AI adoption by delivering integrated Dell and NVIDIA capabilities to accelerate your AI-powered use cases, integrate your data and workflows, and enable you to design your own AI journey for repeatable, scalable outcomes. To learn more, visit: The Dell AI Factory with NVIDIA | Dell USA.