Mr. Tarun Sharma is a PhD candidate at the Department of Electronics and Communication Engineering, Indraprastha Institute of Information Technology, Delhi. His doctoral research focuses on improving the performance and sustainably of heterogeneous CPU-GPU based multicore architecture. He also focuses on customising the open-source RISC-V architecture for various applications.
Tarun holds a bachelor’s degree in electrical and electronics engineering from Guru Tegh Bahadur Institute of Technology, Delhi and a master’s degree in microelectronics from Birla Institute of Technology and Science (BITS) Pilani, Hyderabad Campus.
As a Fulbright-Nehru Doctoral Research fellow, Tarun is aiming to enhance transformer inference using Wireless Network-On-Chip (WNoC) and Machine Learning-based DRAM prefetcher. The proposed framework is expected to improve transformer inference while reducing the power required for off-chip DRAM access, while not requiring any changes in the software and is independent of any CPU, GPU, or accelerator hardware.