Inspirit IoT Awarded Phase II NSF SBIR Grant to Support Development of DNN Architect
Elite US Grant accelerates development of Inspirit’s DNN Architect, an End-to-End Machine Learning Optimization, Architecture Design and Deployment Toolsuite
INSPIRIT IOT, INC., a company developing sensor intelligence and hardware accelerator design tools for the global market, announced today that it has been awarded a National Science Foundation Phase II grant for $750,000 to develop machine learning optimization tools, neural processor optimization tools, and hardware deployment platforms for sensor intelligence. This new grant continues an earlier $225,000 SBIR Phase I grant that developed fundamental technologies to automate optimization, hardware acceleration and deployment of machine learning.
Inspirit’s initial target verticals include Security and Surveillance, Autonomous Driving, Predictive Maintenance, and Healthcare, which represent hundreds of billions USD in growth markets for IoT devices. Wide deployment of these technologies is expected to lead to substantial energy savings and a corresponding reduction in carbon emissions and operational costs.
The award will be used to accelerate development of Inspirit IoT’s DNN Architect, which includes DNN graph and hardware optimizations, Neural Network processing unit design and optimization, and deployment architectures. Together with Inspirit’s prior $600K Seed funding and $2.05M Late-Seed Funding from Senscape Technologies, Inspirit IoT is accelerating development of their team, tools, IoT devices, and commercialization activities.
“The National Science Foundation supports small businesses with the most innovative, cutting-edge ideas that have the potential to become great commercial successes and make huge societal impacts,” said Barry Johnson, Director of Division of Industrial Innovation and Partnerships at NSF. “We hope this seed funding will spark solutions to some of the most important challenges of our time across all areas of science and technology.”
“We are pleased with NSF SBIR’s continuing support of our efforts to develop automated solutions for optimization of neural networks, hardware acceleration architectures, and deployment platforms to enable smart IoT applications,” said Dr. Deming Chen, Co-Founder and President of Inspirit IoT. “The broader impact/commercial potential of this SBIR Phase II project will result in a significant improvement in the performance, power, and cost of deploying machine learning solutions through horizontal platform technologies that enable many vertical applications.”
Dr. Kyle Rupnow, Co-Founder and CTO of Inspirit IoT and principal investigator on this grant stated, “The SBIR Phase II Project focuses on design of high performance, energy-efficient platforms for machine learning applications, and associated design tools and libraries. This improvement will accelerate deployment of intelligent systems and improve scalability through localized intelligence.”
Neural networks are heavily used for machine learning problems, but optimizing for deployment requires extensive trial-and-error and skilled, experienced users to select graph optimizations, hardware implementation optimizations, hardware architecture organizations, and deployment platforms. DNN Architect applies graph optimizations and hardware-friendly optimizations, explores hardware IP selection and architecture organizations, and maps to deployment platforms to achieve high performance and power/energy efficiency goals.
Inspirit IoT is currently in beta with 3 core products: Xcelo Compiler, DNN Optimizer, and SoundGuardian; products already in the hands of the world’s leading chip makers and OEM’s.