How Developers Scale ML at the Embedded Edge
SiMa.ai will begin shipping its SiMa.ai Palette, a software platform for complete ML stack development.
Powered by SiMa.ai’s Machine Learning System on Chip(MLSoC™) silicon, the all-in-one developer platform supports any ML workflow customers plan to deploy on the edge without compromising performance and ease of use. Palette, combined with the availability of two new MLSoC-based production board-level products, are the latest innovations driving the adoption of SiMa.ai’s software-centric, purpose-built MLSoC platform.
Solving ML at the edge requires a heterogeneous compute environment that combines ML, computer vision and application layer programming into a single solution. Developers often face trade-offs in the capabilities of development tools, limited in the algorithms they can support and resulting performance driving the need to resort to extensive hand coding and integration. Lack of scalable tools that can support the process through to deployment is hindering the deployment of ML. The SiMa.ai Palette addresses these problems by combining end to end development tools that are flexible to support any problem while delivering high performance results with pushbutton operation, delivering effortless ML development for deployment and scaling at the edge. The result is a cohesive approach that enhances the developer experience in quickly creating, building and deploying an ML solution.
The SiMa.ai Palette offers single integration through an innovative, simplified and automated development flow that focuses on the developer challenges of creating ML applications. GStreamer workflows using advanced scripting and automation tools combine optimized NN Models, pre-and post-processing functions and applications using those pipelines into a single container for deploying to large numbers of edge devices simultaneously.
“Software is the Achilles heel of ML for the embedded edge. Palette is the software done right, delivering on the promise to developers, simultaneously delivering the flexibility, performance and ease-of-use that we have committed to in our Any network, 10x performance, pushbutton strategy for accelerating edge ML deployment,” said Krishna Rangasayee, CEO and Founder, SiMa.ai. “
The Palette is in alpha with early access customers with full scale deployment starting in Q2 with a license that includes updates, maintenance and support.