(Processing Systems Lab)

Computational LDOs and Autonomous Gain Tracking
The first computationally-controlled Low Dropout Regulator relying on Vdd statistics to auto-tune loop-gain
Learn More
Minimum (Total) Energy Point Tracking under Performance Constraints
Computationally enabled MEP tracking that minimizes total energy dissipation under performance guarantees.
Learn More
Highly-compact delta-encoded ECoG recording
A multiplexed neural-recording front-end that exploits neural-signal statistics to achieve a robust architecture with 10X improvement in neural recording density.
Learn More
Computationally-locked all-digital PLLs
The fastest PVT-robust PLL cold-start and re-lock times for system clocking ever recorded in literature.
Learn More
A unified clock and power regulation architecture for digital SoC domains, virtually eliminating supply noise and temperature-related supply guardbands.
Learn More
Adaptive machine learning
Learning around failures in low-voltage on-chip memories to enable energy-efficient neural network accelerators.
Learn More
Quasi-resonant clocking
The first-ever voltage-scalable, frequency-independent and DVFS-enabled resonant clock architecture.
Learn More

About PSyLab

PsyLab primarily focuses on digital and mixed-signal circuits and architectures for information processing. Recent areas of emphasis include biomedical electronics and energy-efficient computing in current and emerging technologies. We characteristically exploit observations at the system and architecture level to arrive at new, creative solutions to important problems. This common thread is visible across our work on neural (ECoG) recording architectures, computationally locked PLLs, machine-learning acclerators, next-generation clocking architectures, voltage regulation, sensing, and signal processing problems. Identifying the connections that exist across different domains of Electrical and Computer Engineering have allowed us to demonstrate state-of-the-art solutions in these areas.