Chaitali Chakrabarti

Title: Chaitali Chakrabarti: Innovator in Power Management and Deep Learning

Introduction

Chaitali Chakrabarti is a prominent inventor based in Tempe, Arizona, known for her significant contributions to the fields of power management and deep learning. With two patents to her name, she has developed innovative solutions that enhance the efficiency of embedded systems and neural networks.

Latest Patents

Chakrabarti's latest patents include "HiLITE: Hierarchical and Lightweight Imitation Learning for Power Management of Embedded SoCs." This invention addresses the challenges of dynamic power management in modern systems-on-chip (SoCs) by utilizing a hierarchical imitation learning framework. HiLITE maximizes energy efficiency while meeting soft real-time constraints, improving the energy-delay product by an average of 40% and reducing deadline misses by up to 76%.

Another notable patent is focused on "Memory Compression in a Deep Neural Network." This invention involves compressing a fully connected weight matrix in a deep neural network to reduce memory footprint and computational power without sacrificing accuracy. By designating active weight blocks during training, the approach ensures efficient hardware implementation of DNN applications.

Career Highlights

Chakrabarti has worked at esteemed institutions such as Arizona State University and the University of Arizona. Her work has significantly impacted the fields of embedded systems and artificial intelligence, showcasing her expertise and innovative thinking.

Collaborations

Chakrabarti has collaborated with notable colleagues, including Jae-sun Seo and Deepak Kadetotad, further enhancing her research and development efforts.

Conclusion

Chaitali Chakrabarti's contributions to power management and deep learning exemplify her innovative spirit and dedication to advancing technology. Her patents reflect a commitment to improving energy efficiency and computational effectiveness in modern systems.