Chamika SudusingheChamika is a Ph.D. student in the ADAPT@Illinois research group, advised by Prof. Charith Mendis. Chamika is broadly interested in Compilers, Computer Architecture, Deep Learning and Performance Optimization. His current research focuses on building data-driven cost models that guide compiler decisions across heterogeneous hardware platforms and diverse workloads. Chamika has been recognized with several prestigious honors, including the NSF Graduate Research Fellowship (NSF-GRFP), the Lance Stafford Larson Award, the Upsilon Pi Epsilon Honor Society Award, the Richard E. Merwin Scholarship, and the the ICT Student of the Year Award (2022). Before joining UIUC, Chamika conducted collaborative research with the Embedded Systems Lab at the University of Florida, working with Prof. Prabhat Mishra, and with the Centre for Data Analytics and Cognition (CDAC) at La Trobe University, under the guidance of Prof. Damminda Alahakoon. He also worked at WSO2, Nethermind Research, LiveRoom, Nirvana Labs, and Aegis Studio. Chamika is currently excited about a range of problems at the intersection of compilers, machine learning, and hardware systems, including:
Email / Resume / Bio / Google Scholar / Twitter / Github 🌐 Visit Chamika's personal website for more. |
|
Recent News
|
ResearchChamika's research focuses on the intersection of compilers, machine learning, and hardware systems. Key research work include: data-driven cost models, sparse tensor acceleration, machine learning for compiler optimization, network-on-chip security, and hardware-assisted machine learning. A selected representative publications are highlighted. |
|
COGNATE: Acceleration of Sparse Tensor Programs on Emerging Hardware using Transfer Learning
New!
Chamika Sudusinghe, Gerasimos Gerogiannis, Damitha Lenadora, Charles Block, Josep Torrellas, Charith Mendis ICML, 2025 AbstractExploiting feature homogeneity while mitigating hardware heterogeneity to efficiently fine-tune cost models for sparse tensor programs on emerging platforms. |
|
Building Interpretable Predictive Models with Context-aware Evolutionary Learning
Binh Tran, Chamika Sudusinghe, Su Nguyen, Damminda Alahakoon Applied Soft Computing, 2023 (Vol. 132) AbstractBuilding interpretable predictive models by leveraging unsupervised context extraction to improve both accuracy and interpretability on complex datasets. |
|
Eavesdropping Attack Detection using Machine Learning in Network-on-Chip Architectures
Chamika Sudusinghe, Subodha Charles, Sapumal Ahangama, Prabhat Mishra IEEE Design & Test, 2022 AbstractDetecting eavesdropping attacks in network-on-chip architectures, achieving high accuracy with minimal runtime overhead through offline training and collective decision-making. |
|
Denial-of-Service Attack Detection using Machine Learning in Network-on-Chip Architectures
Chamika Sudusinghe, Subodha Charles, Prabhat Mishra NOCS, 2021 AbstractDetecting denial-of-service attacks in network-on-chip architectures, achieving high accuracy with minimal overhead even under diverse and unpredictable traffic conditions. |
|
Hardware-Assisted Malware Detection using Machine Learning
Zhixin Pan, Jennifer Sheldon, Chamika Sudusinghe, Subodha Charles, Prabhat Mishra DATE, 2021 AbstractLeveraging low-level hardware signals such as performance counters and on-chip network traffic, offering faster and more resilient detection than traditional software-based methods. |
|
Network-on-Chip Attack Detection using Machine Learning
Chamika Sudusinghe, Subodha Charles, Prabhat Mishra Springer Book on Network-on-Chip Security and Privacy, 2021 AbstractDetecting denial-of-service attacks in network-on-chip architectures, offering robust detection even under dynamic and unpredictable traffic conditions. |
Talks and Service
|
|
