May 29, 2019 | Atlanta, GA
Two tech giants, Google and Facebook, have each recognized Tushar Krishna with their annual faculty research awards.
Krishna, an assistant professor in the School of Electrical and Computer Engineering at Georgia Tech and affiliate faculty member of the Machine Learning Center at Georgia Tech, is working to address challenges in custom hardware accelerators.
For Facebook, Krishna’s proposal focused on generating custom artificial intelligence hardware platforms that are optimized for target algorithms. The hardware is also optimized for performance or energy constraints on the cloud or edge. His idea is to create a library of lego-like heterogeneous hardware building blocks to run machine learning computations and assemble them via an automated tool-flow.
Krishna was one of eight scholars selected for the Facebook Faculty Research Award, each worth $50,000.
His Google proposal shifted gears and focuses on enabling rapid design and development of machine learning accelerators for the cloud or edge. This grant will allow him to further explore his idea to leverage reinforcement learning algorithms to efficiently represent and search through the hardware design-space of a machine learning accelerator and determine an optimal configuration. He is one of 158 recipients worldwide. Each grant is worth $66,000.
“These awards will allow me to explore a novel research direction of using machine learning to facilitate and automate hardware design, and interact with domain-experts from Google and Facebook. I am grateful for the support from two of the biggest companies in the world,” said Krishna.
The awards go into effect in June 2019.