SM9
SERIES
Pin-to-pin compatible with NVIDIA Jetson modules, designed for high-density edge AI deployment.
Enterprise Computing Performance
SM9 is a high-efficiency 8-core ARM CA53 computing module tailored for 16-channel HD intelligent video analytics at the edge.
The module supports mixed-precision inference, advanced ISP processing, and hardware acceleration for stitching, depth estimation, and dewarp pipelines.
With Jetson pin compatibility and industrial interface density, SM9 enables rapid migration and cost-optimized deployment for AI cameras, robotics, and NVR platforms.
Scale from Endpoint Modules to BM1688 Edge Nodes
Use SM9 for high-density multi-channel vision inference at the endpoint, then pair with BM1688 for higher-compute aggregation, model consolidation, and cross-device orchestration. This topology keeps deployment paths consistent from camera-side modules to edge servers.
Architectural Superiority
High Energy Efficiency
8-core ARM CA53 architecture with INT4/INT8/FP16/BF16/FP32 support for diverse edge AI workloads.
Multichannel Multimedia
Hardware acceleration for 16-channel decode and 10-channel encode, supporting 8K/4K/1080p processing.
Integrated CV Engine
Built-in DPU, stitching, and fisheye dewarp acceleration for real-time computer vision pipelines.
Rich Storage and Memory
Configurations up to 16 GB LPDDR4 and 64 GB eMMC 5.1 for high-throughput, persistent edge inference.
High-Speed I/O Expansion
Dual PCIe Gen3 lanes, dual USB 3.0, dual Gigabit Ethernet, SATA 3.0, and CAN for platform integration.
Industrial Reliability
Compact 260-pin SO-DIMM form factor with optional industrial temperature support down to -40 C.
Technical Specifications
Comprehensive parameter comparison for SM9 series modules.