The edge-side AI based on RK3576 and RK3588 will witness an explosion
The Dingchang RK3576 and RK3588 motherboards will make their appearance in the form of brand-new AIoT in 2025. More large language edge-side models of artificial intelligence and edge computing projects will be supported. In collaboration with Tencent, Baidu, Alibaba, and ByteDance, a large model interface package named DC_LLM_RAG_SDK will be provided, which is suitable for the model transplantation packages of models below 10B of Yuanhun, Wenxin, Tongyi, and Doubao locally.
The high computing power 6TOPS NPU of RK3576 and RK3588 will lay the foundation for deploying large models locally. The DC-LLM-RAG-SDK large model interface package has integrated a vertical industry knowledge base and an update tool. Ecological partners only need to update the text of the knowledge base to use the local large model.
Partners can develop question-and-answer products and generation products with large models + AIGC algorithms, such as intelligent question-answering robots, service navigation robots, intelligent vending machine product recommendations, speech recognition and translation, text-to-image intelligent painting screens, digital humans and other related complete machine products.
With the popularity of DEEPSEEK, the editor of Dingchang predicts that 2025 will be the first year of edge-side AI. Major chip manufacturers and motherboard manufacturers will all lay out edge-side AI products. Dingchang started to lay out the AIoT product line very early. Currently, there are dozens of deployment cases of AI algorithm demos for reference, providing ecological partners with the means to study how to deploy AI algorithms on PC sides and motherboard sides. They can also conduct secondary development for commercial use and quickly implement projects.
If you have any questions, you are welcome to call or add our customer service for further understanding. You are also welcome to leave your project development needs in our online message. We will provide your company with professional technical replies within 24 hours. Let's welcome the new era of AI together.