Purdue accelerates research through adopting novel AI tools, increasing supercomputing prominence

WEST LAFAYETTE, Ind. — Purdue University is solidifying its leadership in advanced research and technology by adopting AI tools and upgrading its high-performance computing (HPC) infrastructure. This forward-looking approach will reinforce Purdue’s research capacity, allowing faculty and students to tackle complex challenges across diverse disciplines.

As AI technologies continue to transform industries and research, Purdue has committed to equipping the university with state-of-the-art computational tools and enhanced HPC infrastructure. This supports AI-driven research across a wide variety of fields, from precision agriculture and environmental health to biomedical research and advanced engineering.

“Embracing AI tools and consistently upgrading our HPC infrastructure are essential steps in maintaining our leadership in cutting-edge research,” said Karen Plaut, Purdue’s executive vice president for research. “These advancements allow our research community to tackle complex problems at an unprecedented scale, from simulating intricate biological systems to developing next-generation materials. By investing in these technologies, we empower our researchers to push the boundaries of knowledge, drive meaningful discoveries, and prepare our institution to meet the evolving demands of science and society.”

Speed for solutions

Purdue research has always been about solving problems for Indiana and beyond. And speed in reaching each solution is as intrinsic to the university’s culture as it has ever been.

In 1963, months after establishing the nation’s first computer science degree, Purdue dedicated its state-of-the-art high-speed IBM 7090 digital computer. As the superfast trailblazer of its time, it had a processing capability of about 229,000 instructions per second.

Today, with speeds measured in petaflops — one quadrillion (1,000,000,000,000,000) operations per second — Purdue’s Anvil supercomputing cluster ranks as the 8th most powerful academic system in the U.S., maintaining an elite status in supercomputing with a peak performance of 5.1 petaflops. And, as a result of $5 million in supplemental funding from the National Science Foundation (NSF), the Anvil supercomputer’s capabilities are being expanded to support AI as part of Purdue’s participation in the National AI Research Resource (NAIRR).

Adding to Purdue’s high-performance computing capabilities, particularly in support of AI, is Purdue’s newest community cluster supercomputer. Launched on Oct. 7, the “Gautschi” supercomputer, named for Walter Gautschi, Purdue professor emeritus of both computer science and mathematics, features cutting-edge technology for traditional simulation and research while also offering NVIDIA’s top-tier solutions for AI applications. It is expected to rank among the top supercomputers worldwide and among U.S. universities when the annual Top 500 list of supercomputers is released in November.

These supercomputers and the AI research scientists supporting their use are critical to the computational research happening all across Purdue, serving over 3,800 research faculty and graduate students across all campuses, colleges and over 66 departments.

New and enhanced computational tools

Purdue continues to adopt and enhance computational tools dedicated to advancing research. Those tools include:

• Microsoft Copilot: Access to chat-based generative AI tools using OpenAI GPT models is available today at copilot.microsoft.com to all Purdue faculty, staff and students.
  
  
• iThenticate 2.0:  An updated tool, outlined in this iThenticate video, is used by researchers, academics and publishers worldwide to detect plagiarism and AI writing.
  
  
• Scite.ai: An innovative research tool recently acquired by Purdue, scite.ai is now available for use by faculty, staff and students through Purdue Libraries. Scite.ai is an AI platform for discovering and evaluating scientific articles. It allows users to see how a publication has been cited and to explore literature quickly and effectively. 
  
  
• Library Guide on Artificial Intelligence: Purdue Libraries recently published a new Library Guide on Artificial Intelligence designed to equip Purdue faculty, staff and students with a range of AI resources carefully selected for their relevance and effectiveness. It serves as a toolbox for research, teaching and learning purposes, and image creation. 
  
  
• Rohan: Purdue began piloting another AI tool, Rohan, with a select group in February 2024 to assist with proposal writing. Rohan is based on domain-aware generative AI. Domain awareness is created by users uploading and managing access to relevant products (e.g., past proposals) that create the basis for generating new materials. Rohan reduces the time needed to find the most relevant material and manage access to key information. Its goal is to increase productivity and quality in the preparation of research proposals.

Advancing Research and AI

Purdue researchers not only use AI, they continually reinvent it. “Purdue researchers are doing AI research and building new large language models,” says Preston Smith, executive director of Purdue’s Rosen Center for Advanced Computing. “Faculty in computer science and electrical and computer engineering are building new models to advance the state of AI itself.”

Purdue’s use of AI has set an aggressive pace for research that intertwines diverse fields that once were siloed. Helping set that pace is IPAI, the Institute for Physical Artificial Intelligence, which is applying AI to manufacturing, aviation, transportation, health care and agriculture. It is developing smart transportation systems, AI-based manufacturing processes and digital forestry, all geared toward real-world impact.

IPAI’s work to bridge the gap between virtual and physical AI applications is one of the four pillars of the Purdue Computes initiative, the university’s comprehensive program that leverages AI technologies across multiple fields and contributes to Purdue’s ongoing standing as a national leader in computing and AI research. The three other pillars are the advancement of computing departments and academic growth, quantum science and engineering, and a strong focus on semiconductors and AI hardware.

A sampling of Purdue research enabled and accelerated by AI includes:

• Working on incorporating human emotion into AI models. This project within IPAI is making discoveries in the new  interdisciplinary field of affective computing — using machine learning and other computer science methods to program artificial intelligence programs to better incorporate and understand human behavior and emotion.
• The prediction of maize yields using AI networks and drones by researchers in the College of Agriculture. This is one of myriad examples of AI in the college, which is a  major contributor to AI-CLIMATE, the AI Institute for Climate-Land Interactions, Mitigation, Adaptation, Tradeoffs and Economy. The research uses AI to develop climate-smart agricultural practices that can help absorb and store carbon while boosting economic sustainability.
• Purdue’s collaboration with Google. Purdue and Google have teamed up for AI research to mitigate climate change. They are using AI to invent low-carbon industrial building design technologies by investigating the viability of new materials, technologies and design strategies. The aim is to significantly reduce the carbon footprint of industrial buildings, such as data centers in the U.S. and around the world.
• A multidisciplinary project using AI to model, simulate and analyze myriad cyber-physical systems, such as military drones, to make them stronger and improve how they’re analyzed. The multiphase $6.5 million project, called FIREFLY, is sponsored by the Defense Advanced Research Projects Agency (DARPA).
• Purdue’s participation among a select group of institutions using AI for cybersecurity. The university is a principal player in the Agent-based Cyber Threat Intelligence and OperatioN (ACTION), funded by the NSF. Joining the University of California, Santa Barbara, and the University of California, Berkeley, in spearheading the five-year research project, Purdue is bringing AI’s continuous learning — and reasoning — capabilities to bear on the constant situational awareness and actions that are fundamental to cybersecurity.

This Purdue AI project compilation provides even more demonstrations of Purdue’s AI use to tackle complex challenges in agriculture, the environment, cybersecurity and human-centered AI.

AI also is an integral component of Purdue’s educational experience. “AI computing resources are a critical capability for us to have on campus so that we’re able to be competitive in doing cutting-edge research and training students with skills that they will need to be able to participate in the workforce In the next 10 to 20 years,” Smith says. “It’s going to touch every bit of their lives. So, for our students, having a good grounding in AI is going to be as critical a skill as computer literacy or a basic understanding of math.”

About Purdue University

Purdue University is a public research institution demonstrating excellence at scale. Ranked among top 10 public universities and with two colleges in the top four in the United States, Purdue discovers and disseminates knowledge with a quality and at a scale second to none. More than 105,000 students study at Purdue across modalities and locations, including nearly 50,000 in person on the West Lafayette campus. Committed to affordability and accessibility, Purdue’s main campus has frozen tuition 13 years in a row. See how Purdue never stops in the persistent pursuit of the next giant leap — including its first comprehensive urban campus in Indianapolis, the Mitch Daniels School of Business, Purdue Computes and the One Health initiative — at https://www.purdue.edu/president/strategic-initiatives.

Source: Preston Smith, psmith@purdue.edu