Research Overview
SHaRe spans various research on electronic circuit and system design and optimization. Our research at the forefront of neuromorphic and bio-inspired computing research. Our research areas and projects include:
- Neuromorphic Sensing & Computing: Neuromorphic sensing and computing integrates brain-inspired sensing modalities with event-driven processing architectures. In these systems, sensors mimic biological sensory organs by converting environmental stimuli—such as changes in light, sound, or touch—into asynchronous spikes. These spikes are then processed in real time by neuromorphic circuits that emulate the dynamics of neuronal networks, enabling rapid, low-power responses and adaptive learning. This approach not only reduces the data bottleneck by transmitting only meaningful changes, but it also paves the way for energy-efficient and highly responsive applications in robotics, autonomous systems, and wearable devices.
- AI-on-a-Chip and Embedded Intelligence: AI-on-a-Chip refers to the integration of artificial intelligence functionalities directly onto microchips. This research focuses on designing specialized processors optimized for neural network computations and machine learning algorithms, with an emphasis on low latency and low power consumption. These chips are crucial for enabling on-device inference and incremental learning in mobile devices, IoT applications, and edge computing systems, thereby reducing the dependency on cloud-based processing.
- Memristor-Based Computing: Memristor-based systems harness the unique properties of memristors—electronic devices whose resistance changes based on their past electrical activity—to emulate the adaptive behavior of biological synapses. These devices enable in-memory computing, allowing data storage and processing to occur in the same location. This leads to compact, energy-efficient architectures ideal for neuromorphic systems and deep learning accelerators, where dynamic, low-power computation is essential.
- Hardware Security and Trust: Our Hardware Security and Trust research focuses on designing and verifying computing systems that are resilient against physical and digital attacks. Researchers in this area develop secure architectures and countermeasures—such as tamper-resistant integrated circuits, hardware-based cryptography, and physically unclonable functions (PUFs)—to protect against side-channel attacks, hardware Trojans, and reverse engineering. By ensuring that both the design and supply chain of hardware components are trustworthy, this field is critical for maintaining data integrity, confidentiality, and system reliability in mission-critical applications ranging from embedded systems to large-scale data centers. Exploring neuromorphic approaches to enhance hardware security and developing efficient implementations of cryptographic algorithms.
- Bio-Inspired Computing Systems: Bio-inspired computing draws on principles from biology—such as adaptation, self-organization, and evolution—to develop novel computational models and algorithms. By studying how natural systems solve complex problems, researchers create algorithms that are inherently robust and efficient. This interdisciplinary approach influences areas such as swarm intelligence, neural networks, and evolutionary algorithms, leading to innovative solutions for optimization, pattern recognition, and adaptive control.