Project Grant W911NF2010003

This Project Grant from the National Science Foundation's $1.2 million Computer and Information Science and Engineering program will fund the development of the Machine Learning Accelerator Cohort Architecture at the University of Southern California from October 2022 to September 2025. The ML Accelerator Cohort Architecture aims to build a heterogeneous computing fabric to efficiently support diverse machine learning models, including novel approaches for privacy-preserving computation. Key...

This Project Grant award from the National Science Foundation's (NSF) Computer and Information Science and Engineering (CISE) program (CFDA 47.070) provides $1,200,000 to Purdue University to develop scalable and ultra-low-power neural accelerators based on 2D ferroelectric semiconductors. The research aims to address hardware needs for future artificial intelligence (AI) platforms by utilizing the unique properties of ferroelectric semiconductors to design energy-efficient circuits and...

This three-year, $290,739 Project Grant from the National Science Foundation's Computer and Information Science and Engineering program will support the development of a novel non-volatile nano-second right-grained reconfigurable architecture for data-intensive machine learning and graph computing applications. The proposed computing architecture, called Right-Grained Reconfigurable Architecture (RGRA), combines aspects of coarse-grained reconfigurable arrays and field-programmable gate arrays...

This $200,000 Project Grant award from the National Science Foundation (NSF) Computer and Information Science and Engineering (CISE) Federal Grant Program (CFDA 47.070) supports the development of a novel chip architecture called CO-FABPRO (Concurrent Forward and Backward Propagation) that aims to revolutionize the training process for long-sequence machine learning models. The project, led by the University of Rhode Island, seeks to make AI training dramatically faster and more energy-efficient...

The National Science Foundation awarded a $599,304 three-year Project Grant to the University of California, Santa Barbara under the Computer and Information Science and Engineering federal grant program (CFDA 47.070). The grant will support the development of novel methods and hardware architectures for optimization and acceleration of spiking neural networks. Key products will include algorithms, software design tools, and field-programmable gate array hardware architectures and...

This $600,000 Project Grant award from the National Science Foundation's Division of Computer and Network Systems will support Purdue University's research to enable real-time, high-bandwidth network traffic analysis using machine learning on programmable switches. The project aims to develop novel methods for efficiently mapping popular ML models like decision trees and neural networks onto programmable switch hardware, as well as new switch primitives for computing flow statistics. This will...

This Project Grant award for $200,000 from the National Science Foundation (NSF) Engineering program (CFDA 47.041) supports research by Southern Illinois University Carbondale (SIU) to develop hardware-constraint-aware design and optimization techniques for memristor-crossbar-array (MCA) based neural network accelerators. The research aims to bridge the gap between software design tools and the physical limitations of MCA hardware, enabling more efficient and cost-effective development of...

This $618,159 National Science Foundation project grant supports research into developing energy-efficient hardware and software for machine learning and artificial intelligence systems. Funded under the Computer and Information Science and Engineering program, the award supports The Washington University in investigating frameworks using quantum-tunneling dynamic-analog memory devices and novel online learning algorithms. Specific objectives include exploring Fowler-Nordheim dynamic analog...

This $256,000 National Science Foundation project grant funds the development of an arithmetic and math hardware accelerator capable of processing compressed floating-point data by Arith Inc. The accelerator aims to dramatically improve computation performance, storage effectiveness, and energy efficiency for computationally intensive applications by enabling arithmetic operations directly on compressed data, rather than requiring decompression first as with current computer systems....

The National Science Foundation awarded a $525,000 Project Grant to the University of California, San Diego under the Computer and Information Science and Engineering program (CFDA 47.070) to support research investigating energy-efficient persistent learning-in-memory with quantum tunneling dynamic synapses from October 1, 2022 to September 30, 2025. The award will fund the development of novel learning hardware and software tools to significantly improve the energy efficiency of artificial...