Stefan Abi-Karam
I am a PhD student at Georgia Tech in the
School of Electrical and Computer Engineering
under my advisor
Prof. Cong (Callie) Hao
as part of
Sharc Lab.
My main research area is in the intersection of hardware acceleration and applied deep learning as well a mix of other topics, including electronic design automation (EDA), FPGA architecture, VLSI algorithms, non-linear optimization, and high-level synthesis (HLS).
I am also a research faculty at Georgia Tech through
Georgia Tech Research Institute (GTRI)
within the
Cybersecurity, Information, Communication, Command and Control and Software Systems (CIPHER)
lab division.
Contact
Email: [email protected]
GitHub: github.com/stefanpie
LinkedIn: www.linkedin.com/in/stefanabikaram
Google Scholar: https://scholar.google.com/citations?user=y1QO9ooAAAAJ
News
-
I got the opportunity to attend the DAC 2023 conference for the first time in San Francisco, CA, where I participated in the SIGDA University Demonstration event. I presented a live FPGA demo running on the Xilinx ZCU102 board for our INR-Arch work. It was also my first visiting San Francisco! I also got to meet up for dinner with our former lab mate, Akshay Karkal Kamath, who now works at Apple, as well as Rishov and Hanqiu, who are interning in the Bay Area over the summer.
[2023-09-13] | Link 1 | Link 2 | -
I attended the HPCA 2023 conference for the first time in Montreal, Canada, with my colleague Rishov Sarkar, who presented our FlowGNN paper. This was my first time visiting Montreal as well as Canada!
[2022-01-01] | Link 1 | -
I started my Ph.D. as a continuation of my Master's work under Prof. Callie Hao at Georgia Tech. Business as usual!
[2023-01-01] -
I completed my Master's degree in Electrical and Computer Engineering at Georgia Tech and walked in the 2022 commencement ceremony! My Master's thesis was an inital version of my "GNNBuilder" work under Prof. Callie Hao.
[2022-12-16] | Link 1 | -
I presented GNNBuilder as part of the 1st Workshop on Democratizing Domain-Specific Accelerators (WDDSA 2022) as part of the MICRO 2022 conference in Chicago, IL. This was my first time attending MICRO and visitng Chicago!
[2022-08-02] | Link 1 | Link 2 |
Projects
-
Graph Neural Network Accelerators
Building FPGA accelerators for graph neural networks (GNNs) using HLS. Also building an automated framework to translate PyTorch models to GNN accelerators. -
Graph Neural Networks for Electonic Design Automation and Hardware Security
Using GNNs to better representations of digital hardware netlists. Also using these representations to build EDA and HOST tools for hardware designers. Currently exploring self-supervised learning and multi-task learning for this problem.
Publications
-
INR-Arch: A Dataflow Architecture and Compiler for Arbitrary-Order Gradient Computations in Implicit Neural Representation Processing
Stefan Abi-Karam*, Rishov Sarkar*, Dejia Xu, Zhiwen Fan, Zhangyang Wang and Cong Hao
[2023-10-29] | Conference: ICCAD 2023 | Link 1 | -
GNNBuilder: An Automated Framework for Generic Graph Neural Network Accelerator Generation, Simulation, and Optimization
Stefan Abi-Karam and Cong Hao
[2023-09-04] | Conference: FPL 2023 | Link 1 | -
M5: Multi-modal Multi-Task Model Mapping on Multi-FPGA with Accelerator Configuration Search
Akshay Karkal Kamath, Stefan Abi-Karam, Ashwin Bhat and Cong Hao
[2023-04-17] | Conference: DATE 2023 | Link 1 | -
FlowGNN: A Dataflow Architecture for Real-Time Workload-Agnostic Graph Neural Network Inference
Rishov Sarkar, Stefan Abi-Karam, Yuqi He, Lakshmi Sathidevi, Cong Hao
[2023-02-25] | Conference: HPCA 2023 | Link 1 | Link 2 | -
GenGNN: A Generic FPGA Framework for Graph Neural Network Acceleration
Stefan Abi-Karam*, Yuqi He*, Rishov Sarkar*, Lakshmi Sathidevi, Zihang Qiao, Cong Hao
[2022-01-20] | Preprint: Arxiv | Link 1 |
Teaching
-
Advanced Computer Architecture - ECE 6100 / ECE 4100 / CS 4290 / CS 6290
Guest Lecturer: Two guest lectures on Network-on-Chip (NoC) fundamentals, architectures, and applications.
[Fall 2023] | Georgia Institute of Technology | Link 1 | -
Parallel Programming for FPGAs - ECE 8893
Guest Lecturer: Guest lecture on real-world applications of high-level synthesis (HLS) for FPGAs, including deep learning.
[Spring 2023] | Georgia Institute of Technology | Link 1 | -
Scientific Computing
Instructor: Teaching high school students Python and a wide range of scientific computing topics to be applied in their own research projects.
[Summer 2020] [Summer 2021] [Summer 2022] | American Heritage School |
Honors and Awards
-
Recipient of the Georgia Tech "Warren Batts and Austin Brown Innovation Award" VIP Scholarship
[2020-03-30] | Georgia Tech | -
3rd Place Oral Presentation
[2019-04-09] | The Georgia Tech Undergraduate Research Symposium |
Service
- Member | Institute of Electrical and Electronics Engineers (IEEE)
- Member | Association for Computing Machinery (ACM)