Neurala, a Boston-based software company that makes robots, drones, consumer electronics, toys and smart devices more intelligent, announced today that it has joined the NVIDIA (NASDAQ: NVDA) Jetson Embedded Platform ecosystem. As a part of the ecosystem, Neurala will continue to advance GPU-accelerated applications of its artificial intelligence technology.
Key to the success of Neurala’s implementation on the NVIDIA Jetson platform is the world’s first algorithm for “instantaneous deep learning.” This patent-pending algorithm enables learning on the device without any internet connection or cloud-based processing. Therefore, it can be used at the edge, where internet connections are impractical, too costly or too time consuming.
Neurala’s pioneering work using NVIDIA GPUs dates back to 2006 when it was the first to see that GPUs could be used as powerful parallel processors for handling input from multiple sensors and that the input needed to be synchronized for real-time decision making. The company has related patents and has extended the work on artificial neural networks (ANN), including deep learning networks, to enable real-time control of autonomous machines. (US Patents 9,189,828 and 8,648,867 for Graphic Processor Based Accelerator System and Method.)
“We have loved working on the NVIDIA platform over the past 10 years. It provides the power, flexibility and capabilities we need in a small, power-efficient package,” said Massimiliano Versace, CEO of Neurala. “We expect that our ongoing work together on both the business and technical sides will enable Neurala to reach more customers and create more applications to meet customer needs.”
“The team at Neurala has done tremendous work in GPU-accelerated artificial intelligence,” said Deepu Talla, vice president and general manager of Autonomous Machines at NVIDIA. “By bringing these capabilities to the Jetson ecosystem, they’ll enable new products and research in autonomous machines.”
Companies are invited to join the NVIDIA Jetson Embedded Platform ecosystem program based on meeting exacting specifications of the Jetson Embedded Platform.