Ahmedullah Aziz's research focuses on moving beyond the physical limits of traditional silicon-based semiconductors to address the core challenges of energy efficiency and performance in the post-Moore's Law era. He integrates device physics with circuit-system co-design, achieving a series of advances in cutting-edge fields such as cryogenic electronics, neuromorphic computing, and post-CMOS technologies.
Aziz and his team are developing computing hardware for extreme environments. In cryogenic electronics, they proposed an innovative superconducting logic family that integrates a heater-cryotron with a ferroelectric superconducting quantum interference device (SQUID) to achieve voltage-controlled logic gates. This technology merges the design flexibility of traditional CMOS with the energy efficiency of superconducting circuits, providing a new path for developing scalable, high-performance cryogenic computing systems.
In artificial intelligence hardware, Aziz explores the physical foundations for future intelligent computing. He and his team developed an artificial synapse based on a superconducting memristor and used it to construct an energy-efficient cryogenic spiking neural network. This research opens possibilities for building low-power, brain-inspired computing systems for specialized environments. Furthermore, his team designed a "smart image sensor" with in-pixel computing, which embeds computational capabilities directly within the pixel to enable real-time image processing. This technology holds significant potential for resource-constrained applications like autonomous navigation and machine vision.
To accelerate the development cycle for novel devices, Aziz applies machine learning to electronic design automation (EDA), developing a framework that can integrate atomistic material physics into compact device models. His goal is to continue exploring emerging materials and unconventional computing architectures to drive the development of energy-efficient, intelligent, and scalable computing hardware.
