Research groups

Applications often expose configurable knobs that impact performance, quality, and resource usage. Achieving optimal configurations is challenging due to conflicting requirements and dynamic system conditions. This research focuses on enhancing applications with self-optimization capabilities for adaptive performance.

The compiler technology group focuses on enhancing compilers for security, emerging architectures, parallel programming, and reverse-engineering. Their work includes detecting and mitigating side-channel leaks, optimizing for new architectures, supporting parallel models, and translating machine code for analysis or modification.

The run-time resource management group addresses challenges in task scheduling and resource mapping for complex, power-hungry architectures. Their research spans from embedded systems to HPC, aiming to optimize system utilization while minimizing power, temperature, and resource contention, supporting the Barbeque Open-Source Project.

The security group focuses on applied cryptography, including efficient implementations for low-power devices, countering side-channel attacks, and cryptography for emerging domains like cyber-physical security, data privacy, and social network security. They also work on secure email and authentication systems.

This research group focuses on computer architecture, specifically for multi-core processors. Their work includes energy-performance optimization, secure architecture design to counter side-channel attacks, and power-efficient embedded multi-core designs for low-power embedded systems.

The group conducts research at the intersection of hardware and software, addressing challenges in both thermal management and wireless sensor networks. Their work includes the design of advanced heatsinks and the development of Dynamic Thermal Management (DTM) software to regulate CPU temperature and ensure system reliability. In parallel, they focus on wireless sensor networks, creating precise time synchronization systems to enhance data transmission and fusion. Through hardware-software codesign, they build platforms that balance power efficiency, performance, and programmability—leveraging the Miosix OS and WandStem nodes for seamless integration.