Shulou(Shulou.com)11/24 Report--

CTOnews.com, November 9 (Xinhua)-- Google has launched an open source framework called Open Se Cura, which aims to accelerate the development of low-power AI systems for "small devices." The framework contains a set of open source tools that are said to accelerate the development of machine learning through "collaborative design and development".

Google pointed out that the speed of contemporary hardware development lags behind the speed of machine learning models and software development, thus causing obstacles to the development of large-scale AI models. At present, consumers' demand for small devices is growing rapidly, but running related models on small devices requires a lot of power, so it is necessary to develop low-power AI systems for "small devices".

The Open Se Cura project currently launched by Google claims to be able to promote cooperation between hardware designers and software developers, promote the concept of "co-design of hardware and software", and maximize the operating performance of AI devices from the bottom of software and hardware. With this project, Google is expected to launch a small AI model for Pixel devices and use its own Tensor chip to run AI assistants with low power on the device side.

CTOnews.com noted that thanks to Open Se Cura's plans to adopt the open source model, any developer interested in the industry can join the project to help improve development tools and provide more creative ideas.

Google claims that as Project Open Se Cura continues to grow, Google will seek more partners. Google is working with universities such as Cambridge University and the University of Michigan to promote open source development programs, which Google claims to have made some achievements in the research on the power consumption of generative AI.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.