CORE-6490JD4 QCS6490 AI System on Module

Powered by the Qualcomm QCS6490, the module integrates the sixth-generation Qualcomm AI Engine and a Hexagon DSP tensor accelerator, delivering up to 12 TOPS. It efficiently handles image recognition, object detection, video analytics, and real-time inference. Hardware-accelerated AI enables reliable deployment of computer vision algorithms, on-device large models, and generative AI while reducing cloud dependence, lowering latency, and protecting edge data.

Built on the Qualcomm Linux software stack with native Ubuntu support, the module enables private deployment of Transformer-based multimodal models, including Gemma, Llama, Qwen, and Phi. It includes the QNN inference framework and supports TensorFlow, PyTorch, ONNX, Keras, and Caffe for efficient model conversion, validation, and end-to-end on-device AI deployment.