Prof. Dr. Nirmala Shenoy

Abstract

The next breakthrough in artificial intelligence will be driven not only by advances in models and accelerators, but by innovations in communication networks. As AI systems scale from thousands to millions of distributed accelerators, communication is rapidly replacing computation as the primary bottleneck.

AI systems demand massive data movement demand networking architectures that provide unprecedented levels of bandwidth, ultra-low latency, resilience, security, and energy efficiency. The network is becoming an integral part of the AI system. Today's communication infrastructures face serious challenges in scalability, reliability, security, energy consumption, operational complexity, maintenance and management as we continue to use retrofitted technologies and solutions designed for communications and technologies that existed several decades ago. These limitations threaten to constrain the future of large-scale AI training, distributed AI inference, and emerging autonomous AI ecosystems. Addressing these challenges requires rethinking networking from first principles and developing architectures and techniques suited to AI-native traffic models and demands and workloads and technologies with a capability to  evolve with trends and demands.

This presentation explores the evolving AI ecosystem and the multidisciplinary research challenges spanning compute, memory, storage, networking, scheduling, resilience, software, cybersecurity, and sustainable infrastructure. Emphasis is placed on networking as a foundational enabler of future AI systems and on the emerging research challenges in AI-native network architectures, scalable routing and multipath forwarding, high-performance transport protocols, network telemetry and AI-driven control, scheduler-network co-design, resilient communication and trustworthy AI networks.

Recognizing this critical need, the IEEE Rochester Joint AESS/Communications Society Chapter has established a Special Interest Group (SIG) on Networking for AI Infrastructure to bring together researchers, industry leaders, and standards organizations to advance the next generation of AI communication systems. The SIG aims to foster collaboration, accelerate innovation, and participate in standards to shape the future of communication infrastructures not limited to the next generation of AI.

Biography:

Short Bio

Dr.-Ing. Nirmala Shenoy is a Professor in the School of Information (iSchool), Golisano College of Computing and Information Sciences at the Rochester Institute of Technology (RIT), where she directs the Networking and Security Laboratory. She earned her bachelor’s and master’s degrees in engineering from the University of Madras, India, and was awarded the prestigious Deutscher Akademischer Austauschdienst (DAAD) Fellowship and pursued her Ph.D. in Computer Science at the University of Bremen, Germany, specializing in network protocols.

Dr. Shenoy began her research career as a Senior Research Scientist and Assistant Director at the Central Electronics Engineering Research Institute (CEERI), part of the Council of Scientific and Industrial Research (CSIR), India. She subsequently pursued teaching and research at the Information Communication Institute of Singapore, established through a collaboration between AT&T Bell Laboratories and Singapore’s National Computer Board. She later held academic positions at several Australian universities before joining RIT in 2001.

Her research focuses on network architectures and communication protocols for enterprise, data center, wireless, ad hoc, and sensor networks. A central theme of her work is understanding and reducing the complexity of modern network implementations through clean-slate architectural approaches and innovative protocol designs aligned with emerging networking technologies and requirements.

Dr. Shenoy chaired the IEEE 1910.1 Working Group, which developed the IEEE standard on Meshed Tree Bridging for loop avoidance in switched networks. Her research has received funding from the U.S. Department of Defense, the National Science Foundation, New York State, Boeing, Cisco, and other organizations.