My research sits at the intersection of computer architecture and hardware security. I work on memory encryption and the secure processor stack, studying how Trusted Execution Environments protect confidential workloads on modern cloud hardware, and where current designs fall short. I use gem5 for microarchitectural evaluation.
- ▶Hardware security in Trusted Execution Environments (TEEs)
- ▶Memory encryption and secure processor architecture
- ▶Confidential computing on cloud and server-scale systems
- ▶Microarchitectural security and side-channel resilience
- ▶Computer architecture simulation (gem5)
Derrick Greenspan, Naveed Ul Mustafa, Andres Delgado, Connor Bramham, Christopher Prats, Samu Wallace, Mark Heinrich, Yan Solihin
IEEE SEED 2024 · 2024
Persistent memory objects (PMOs) require protection mechanisms to ensure data integrity at rest. This work proposes LOaPP, a low-overhead approach to at-rest PMO protection that improves performance while maintaining security guarantees for persistent memory workloads.
Currently focused on memory encryption schemes for secure processors, analyzing threat models, designing defenses, and evaluating real-world performance impact using gem5 on Azure across standard benchmark suites. Work targets deployment on server-scale TEE hardware.