My name is Arash (Persian: آرش) (aka NEO). I am a PhD candidate at the School of Computer and Communication Sciences, EPFL, and a doctoral research assistant at PARSA under supervision of Prof. Babak Falsafi. I did my master’s in Software Engineering at the Department of CSE and IT of Shiraz University, where I also received my B.Sc. in Software Engineering.
I am broadly interested in the field of computer systems and interdisciplinary systems-level problems found in modern, large-scale datacenters, from cloud services to data stores and all the way down to server systems. By characterizing modern datacenter applications, I look for maximizing the compute density of server systems and minimizing their energy footprint through specialization of various system components and better system integration. My current focus is on the evasion of the RPC tax in datacenters through hardware-software co-design.
I am currently on the job market. I expect to graduate at the end of the Summer.
You can find my CV (updated July 2021) here.
Ph.D. in Computer & Communication Sciences
2015 - 2021 (Expected)
Doctoral research assistant at the Parallel Systems Architecture (PARSA) lab under supervision of Prof. Babak Falsafi.
Thesis: Hardware-software Co-design to Evade the RPC Tax in Datacenters
Related Courses: Advanced Multiprocessor Architecture, Topics on Datacenter Design, Understanding Datacenter Software Dynamics
M.Sc. in Computer Engineering (Software Engineering)
2013 - 2015
Thesis: Design & Implementation of a Scheme for Big Data Processing on GPU
Advisor: Dr. Farshad Khunjush
Ranked First: Achieving the highest course GPA (19.84/20) among all M.Sc. students
Related Courses: Advanced OS, Advanced Computer Architecture, Multicore Programming, Parallel Algorithms, Grid Computing, Software Architecture, Text Mining
B.Sc. in Computer Engineering (Software Engineering)
2009 - 2013
Ranked First: Achieved the highest GPA in CS courses (18.88/20) among all B.Sc. students
Cerebros: Evading the RPC Tax in Datacenters. Arash Pourhabibi, Mark Sutherland, Alexandros Daglis, Babak Falsafi. To appear in Proceedings of the 54th IEEE/ACM International Symposium on Microarchitecture, MICRO'21, Athens, Greece, October 2021.
Equinox: Training (for Free) on a Custom Inference Accelerator. Mario Drumond, Louis Coulon, Arash Pourhabibi, Ahmet Yüzügüler, Babak Falsafi, Martin Jaggi. To appear in Proceedings of the 54th IEEE/ACM International Symposium on Microarchitecture, MICRO'21, Athens, Greece, October 2021.
Data Transformer Apparatus. Arash Pourhabibi, Siddharth Gupta, Hussein Kassir, Mark Sutherland, Zilu Tian, Mario Drumond, Babak Falsafi, Christoph Koch. Patent (Pending), March 2021. [Detailed Record]
Optimus Prime: Accelerating Data Transformation in Servers. Arash Pourhabibi, Siddharth Gupta, Hussein Kassir, Mark Sutherland, Zilu Tian, Mario Drumond, Babak Falsafi, Christoph Koch. In Proceedings of the 25th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, ASPLOS'20, Lausanne, Switzerland, March 2020. [Detailed Record] [View at Publisher]
Towards Near-Threshold Server Processors. A. Pahlevan, J. Picorel, A. Pourhabibi, D. Rossi, M. Zapater, A. Bartolini, P. G. Del Valle, D. Atienza, L. Benini, and B. Falsafi. In Proceedings of the 2016 Conference on Design, Automation & Test in Europe (DATE), Dresden, Germany, March 2016. [Detailed Record] [View at Publisher]