Low-Power Integrated Circuits and Embedded Systems Laboratory (LPiNS-Lab)
We model and develop low-power algorithms and methodologies to design energy-efficient integrated circuits and computer hardware architectures for machine-learning applications. Foundational and applied research work extends across diverse audiences and applications, from biomedical device innovation to re-designing in-memory device architecture. In short, in our laboratory we are:
- Developing low-power integrated circuits for next-generation biomedical instrumentations such as inkjet-printed AI-core-based respiratory bracelets and in-memory image sensors with ReRAM technology.
- Designing power-efficient RTL-level domain-specific architectures (DSAs) to execute high computational workloads and massively parallel computing.
- Modeling energy-efficient deep learning algorithms to tackle the current AI training and testing bottleneck.
We are a group of intrigued individuals and low-key coffee connoisseurs who love to get their hands dirty!
Research Topics
LPiNS-Lab focuses on solving the energy-efficiency problem in embedded systems and integrated circuits. We develop methodologies for low-power integrated circuits (IC) and hardware modeling for computer processors. With a background in computing, integrated circuits, and embedded systems, we work designing energy-efficient artificial intelligence (AI) accelerators and memory using analog and digital domains. Our laboratory thrives on innovation and envisions a future where technology is a force for good. We are driven by the dual pursuit of societal advancement and environmental sustainability, ensuring that our research has a lasting impact on our communities and planet. Visit our research topics to contribute to our mission and vision.