MemryX Inc. of Ann Arbor, a startup focused on accelerating artificial intelligence (AI) processing for edge devices (any piece of hardware that controls data flow at the boundary between two networks), announced it has formed a partnership with Nota.ai, an independent AI software vendor, to streamline the development, testing, and deployment of AI models on MemryX’s advanced AI accelerator chips.
MemryX’s edge AI chips have broad application in the transportation, Internet of Things (IoT), robotic, industrial, and edge computing industries. They feature high-performance and low-latency AI processing using a proprietary at-memory computing and dataflow architecture.
Developers can now use Nota.ai’s software platform to search, compress, and automatically deploy artificial intelligence models on MemryX edge AI processors.
The partnership was created to help AI developers to quickly port models to MemryX advanced AI chips slashing deployment times from weeks to days. This is meant to allow developers to rapidly design, validate, and test new AI algorithms on physical hardware when using MemryX’s unique architecture.
“Nota.ai and MemryX’s products are complementary, effortlessly resolving AI engineers’ development and deployment pain points,” noted Keith Kressin, president and CEO of MemryX. “Nota.ai’s optimized and compressed AI models targeted to run on our AI accelerators deliver a cost-effective and highly-performant solution. We’re thrilled to partner with Nota.ai because they have a proven solution stack, and together, we can enable the next generation of edge AI solutions.”
The relationship between Nota.ai’s software stack and MemryX’s hardware is meant to open the door to rapid deployment of advanced AI solutions. Among the first is Nota.ai’s Driver Management System (Nota DMS), a lightweight off-the-shelf AI-based driver monitoring system that can substantially improve road safety.
Available to vehicle manufacturers, the Nota DMS solution encompasses driver identification, drowsiness detection, distraction detection, and unsafe driver behavior detection. NetsPresso’s hardware-aware AI model optimization and compression platform allow the embedded AI models in Nota DMS to run unmodified and fully optimized on MemryX hardware.
MemryX AI accelerators support all popular AI software frameworks, including PyTorch, ONNX, TensorFlow and Keras; all popular processors, including x86, Arm and RISC-V; and all modern operating systems, such as Linux, Windows, and Android. Hundreds of AI models have been verified on MemryX silicon, which is sampling now to numerous customers with production planned Q4’23.
“Our partnership with MemryX allows AI developers to move from development to deployment more quickly,” says Myungsu Chae, CEO of Nota.ai. “And enables new features to be quickly integrated into their AI stack. MemryX’s hardware platform delivers the advanced hardware capabilities our customers want, and we expect to see new AI solutions brought to market quickly due to our collaboration.”
