TLDR: Professor Ajay Joshi of Boston University is part of a collaborative team that has secured a $1.5 million NSF grant to develop next-generation, energy-efficient AI chips. This three-year project, co-led by the University of Delaware and involving UC San Diego, aims to integrate photonics (light) with electronics to create faster, more sustainable computing architectures for advanced artificial intelligence, addressing the high energy consumption of current AI hardware.
A groundbreaking three-year, $1.5 million research project funded by the National Science Foundation (NSF) is underway to revolutionize artificial intelligence (AI) hardware by developing light-powered chips. This initiative, part of the NSF’s Addressing Systems Challenges through Engineering Teams (ASCENT) program, brings together leading researchers from the University of Delaware, the University of California San Diego (UCSD), and Boston University (BU).
Professor Ajay Joshi from Boston University is a key collaborator in this effort, focusing on creating the essential computing architecture for these innovative photonic-electronic chips. The project is co-led by Vishal Saxena, a professor in the Department of Electrical and Computer Engineering at the University of Delaware, who will primarily oversee the electronic chip components. Shayan Mookherjea, a professor at UCSD, will be responsible for developing the photonics components.
The core objective of this research is to tackle the escalating energy demands of modern AI. Current AI systems rely heavily on computer chips originally designed for graphics processing, which consume vast amounts of energy. These chips operate by converting information into digital 0s and 1s, requiring billions of tiny switches to flip on and off, each consuming energy. As AI applications become more complex and widespread, from customer service chatbots to advanced medical diagnostics, the sustainability concerns associated with their power consumption are growing.
Professor Saxena highlighted the urgency of this challenge, stating, “Meeting the growing demands of AI requires new chip technologies that are faster and more energy-efficient. Our research explores using light alongside electronics to boost speed, reduce processing delays and lower energy consumption.” He further elaborated on the computational process, explaining, “Fundamentally, what these chips do is multiply two numbers and add them, in massively parallel fashion—trillions of computations in a second. By shifting much of this bulk computation into the analog domain, in which information is handled as continuous signals instead of being translated into 0s and 1s, we can do the same math with far fewer switches, which can save a significant amount of energy.”
The collaborative team will leverage silicon photonics, a technology that utilizes light to process and transmit information directly on a chip, working in conjunction with traditional electronic circuits. Photonics are particularly efficient for the multiplication step in computations, while analog electronics are better suited for the addition step. This hybrid approach also capitalizes on photonics’ ability to carry immense amounts of information simultaneously, facilitating parallel computations.
Also Read:
- Tech Giants Advance AI and Crypto: Coinbase Invests in Indian Exchange, Intel Pursues Chip Business Deal Amidst Broader Small Business Tech Adoption
- Jeff Bezos Backs Seven AI Startups: A Look at His Latest Investments
The $1.5 million grant will be equally distributed among the participating institutions, supporting parallel research efforts. Following the design phase, the team plans to fabricate the new photonic-electronic system and rigorously evaluate its accuracy and energy efficiency using cutting-edge AI models. Beyond the technological advancements, the project is also committed to fostering the next generation of innovators by training undergraduate and graduate students in advanced chip design and AI hardware development.
