Soroush Bateni


I am currently a PhD candidate in the Department of Computer science at University of Texas at Dallas. I started my PhD studies in Spring 2018. My advisor is Prof. Cong Liu. I recieved my Master's degree in Computer Science from UTD in 2018.

My main research area is Cyber-Physical Systems, specifically, GPU-enabled CPS with a specialty in DNN-driven autonomous driving. I have also worked on improving reliablity of General Purpose GPU computing through better scheduling, better system software, and better performance/energy modeling.

My most recent work concerns data-driven resource management on autonomous embedded systems (RTSS 2019) and GPU-dependent Deep Neural Networks and their deployment on resource-restricted embedded hardware (RTSS 2018, RTSS 2018). Prior to that, I helped publish noumerous papers at venues such as RTSS, RTAS, and TDPS.

ECSS 4.215, UT Dallas
Richardson, Texas
Email: soroush AT


  • Our paper titled "Predictable Data-driven Resource Management: an Implementation using Autoware on Autonomous Platforms" has been accepted to RTSS.
  • Finished my Research Internship at Fujitsu Laboratories of America.
  • Two first-author papers have been accepted to the 39th RTSS (RTSS 2018): ApNet (exploring approximation potential of DNN for timing predictability), and PredJoule (a timing-predictable energy optimization framework for DNN-driven autonomous systems).
  • Our paper "S^3DNN: Supervised Streaming and Scheduling for GPU-accelerated Real-Time DNN Workloads" received the Best Paper Award at RTAS 2018 (CPSWeek@Porto).
  • Our paper on "Exploring Computation and Data Redundancy via Partial GPU Computing Result Reuse" has been accepted to the 32nd ACM International Conference on Supercomputing (ICS 2018).
  • NSF ACM/IEEE RTSS Travel Grant 2018