A new AI algorithm platform for its AI BOX

Dingchang's AI BOX and supporting boardsnow simultaneously support AI deployment methods for Android 12 and Ubuntu20.04 systems, including board-side and PC-side deployment.

New featuresinclude board-side Python inference, support for USB cameras, MP4 videos, andOpenCV inference, as well as support for large language AI models such as QWEN,LLAMA, TINYLLAMA, PHI3, CHATGLM3-6B, GEMMA, INTERNLM2, MINICPM, etc.

For the Visual Cabinet development board(DC_R566), a new GOODS_REC multi-category commodity recognition function hasbeen added, integrating dataset creation, annotation, training, board-side inference, and other methods, supporting Android 11 system, among more functions.

The AI BOX and supporting boards（DC_A588） are equipped with the high-end RK3588 main control chip processor, an octa-core 2.4GHz high-frequency CPU, a 6TOPS NPU AI neural processing unit, as well aspractical functions such as SATA hard drive expansion and CAN industrial protocol support, ensuring stronger performance.

The algorithm platform's API documentationis continuously updated to keep up with market demands, providing corporateclients with an enterprise-level AI algorithm deep learning platform thatsupports commercial use, secondary development, systematic learning of AIalgorithm deployment methods, and the use of RK hardware's NPU. It is thepreferred choice for companies with strong R&D capabilities seeking boardsand AI algorithm deep learning suppliers.

Dingchang's AI algorithm engineers providefull guidance on AI algorithm deployment methods and project implementation.New and existing customers are welcome to experience our worry-free services.