My research focuses on the design and development of compute subsystems for autonomous machines. The recent advancements in machine learning have started a revolution with autonomous cars at the forefront. This revolution will lead to intelligent machines with the capability to be our assistant and our companions. However, in order to reach this level of intelligence, cyber physical machines need to be equipped with compute subsystems that are both powerful and energy efficient. This enables them to process their dynamic and unpredictable environments in real time. I focus on optimization techniques across several compute stack layers including application, OS, runtime and architecture, in order to achieve this balance of performance and energy efficiency.
MAVBench Hardware in the Loop Simulator