Qualcomm Innovation Fellowship finalists, Columbia Mishra and Prabhakar Marepalli. Both are Ph.D. students advised by Dr. Jayathi Murthy in the Mechanical Engineering Department at The University of Texas at Austin.
Two Mechanical Engineering Ph.D. students of Chair Jayathi Murthy, Prabhakar Marepalli and Columbia Mishra, have been selected as finalists for the highly selective $100,000 United States Qualcomm Innovation Fellowship (QInF), along with three other teams from The University of Texas at Austin. Marepalli and Mishra will attend the East Coast Finals in Bridgewater, New Jersey on March 24. Their innovation title is "Simulation of Highly Non-Equilibrium Electro-Thermal Transport in Ultra-Scaled Microelectronics." This is the first year that UT Austin students have been able to apply for the fellowships.
Qualcomm engineers carefully reviewed the 137 eligible applications from 18 participating schools and selected 34 as finalist teams, who will present their proposals to a panel of executive judges. The winning teams will be mentored by Qualcomm researchers to facilitate close collaboration and interaction with the company's research and development group. The finalists will be announced this summer. Eight of these proposals are expected to be funded.
Their Project: Modeling Electro-thermal Transport in Semiconductor Systems
As the electronic devices evolve rapidly to perform faster and smarter, there is a growing need for better understanding of thermal transport in semiconductor devices. The heating of a semiconductor device occurs mainly through a process called Joule heating. As the electrons flow in the device channel, they have to overcome the resistance to their transport, and in the process, release energy in the form of heat called Joule heating. When we look at the effect of this heat on a chip with more than a billion transistors, the on-chip power density transcends to an extremely large value. To put this in perspective, if these power densities continue with the same trend, they would be soon comparable to those that we find in rocket nozzles.
Qualcomm demonstrates this in a company video in which they melt butter on three different cell phones as they are turned on.
Project Goal
The goal of our project is to develop a rigorous framework to model and simulate electro-thermal transport in these devices. Traditional thermal modeling approaches fail in transistors, which can be as small as 10 nanometers. In semiconductor devices, heat is primarily carried by particles called phonons, which are quantized lattice vibrations. The transport of phonons can be modeled by the Boltzmann Transport Equation (BTE). Our group already has simulation machinery and unique acceleration algorithms for efficiently solving the Boltzmann transport equation. In this project, we will integrate into this machinery the effects of electron transport in order to predict hot spots in a real semiconductor device. These predictions are expected to provide insight into the device heating phenomenon and its correlation with electrical transport parameters. This understanding can be used to design better consumer devices such smart glasses and faster tablet computers.
