Lab
I will lead a research group in the Department of Electrical and Computer Engineering at UNC Charlotte, starting Fall 2026, working in the areas of computer architecture, systems, and security.
Modern computing is built on the assumption that hardware keeps secrets. That assumption is breaking down: processors share microarchitecture resources, and hardware-based attacks have exploded in the past decade, leading to a steady stream of real-world vulnerabilities. Retrofitted mitigations routinely cost double-digit performance, and emerging workloads such as AI make the problem more severe, as models and the data they touch become the most valuable assets on the machine. Our research aims to make security a first-order design objective: understanding how hardware leaks sensitive information, and re-architecting processors, memory systems, and AI accelerators to be secure by design without sacrificing performance.
To apply, email [email protected] with your CV, transcripts, and a few sentences on your background and which research directions below interest you. Current UNC Charlotte undergraduate and master’s students interested in hands-on research are welcome to reach out.
- Computer architecture/hardware securityModern processors leak secrets through timing and contention in shared microarchitecture. The main idea of this direction is discovering and characterizing new side and covert channels and designing low-cost defenses that prevent leakage from its source.
- Hardware/software infrastructure for secure and privacy-preserving AI/MLML models are prime intellectual property. However, the impact of hardware-level threats on AI/ML workloads is not fully understood yet. The goal of this direction is to understand hardware-level threats to ML systems and build hardware/software co-designs that create end-to-end leakage-resistant stacks across GPUs and multi-tenant NPUs, and support for practical secure inference deployment.
- Secure processor architecturesSecure processor architectures are built to prevent physical attacks in computing systems. However, the holistic impact of secure processor design is underexplored. This direction will focus on investigating new threats introduced by architectural support for secure processors and designing future secure-by-design processor architectures that are secure against all variants of adversarial attacks.
- Security of emerging memory and storageEmerging non-volatile memories introduce new device behaviors that adversaries can exploit. The goal of this direction is to characterize side channels rooted in device physics and to architect secure-by-design integration for high-performance memory and storage.