Multi-GPU Modeling and Performance Improvement

Modeling and evaluating data compression algorithms to improve bandwidth and performance of multi-GPU systems

Architectural Support for Designing Dependable Non-volatile Main Memories

Architectural support to enhance energy-efficiency and durability of encrypted non-volatile main memories

Improving endurance of non-volatile memories using error recovery schemes

Write disturbance errors mitigation in Phase Change Memory using compression and coding techniques

Energy Efficient Computer Systems

Dynamic heterogeneous architectures with online phase predication to improve energy-efficiency of computing systems

Application mapping on many-core systems under process variation to improve reliability and energy consumption

Energy efficient on-chip power delivery with static and dynamic voltage regulator clustering