Quectel Unveils SP900BF-AP AI Smart Module
The Qualcomm-based module targets on-device AI for industrial and autonomous mobile robot applications.
Updated on Nov. 11, 2026 in Semiconductors

Quectel Wireless Solutions has announced the SP900BF-AP smart module, built on the Qualcomm Dragonwing IQ-9075 Processor and delivering up to 100 TOPS of on-device AI compute. Engineering samples are available to selected customers.
Why it matters
Local processing reduces latency, bandwidth costs and data privacy risks associated with cloud-based processing. The module's extended lifecycle commitment targets industrial deployments that require long-term hardware availability.
The module pairs an octa-core Qualcomm Kryo CPU running at up to 2.36 GHz with an Adreno 663 GPU, Hexagon DSP and dedicated NPU, supporting PCIe, USB 3.1, UART, SPI and I²C interfaces. Pricing and full production volumes have not been disclosed, and Ubuntu and Yocto Linux support remains under development.
The players
Quectel Wireless Solutions
A Shanghai-based IoT module manufacturer known for cellular and wireless modules used across industrial and automotive applications.
Qualcomm
A San Diego-based semiconductor company whose Dragonwing IQ-9075 Processor serves as the foundation of the SP900BF-AP module.
The details
The SP900BF-AP processes data locally and supports four-camera input, 3D perception and on-device navigation for autonomous mobile robots. The module has completed key technical validation, and Quectel commits to a product lifecycle of up to 10 years.
Timeline
November 11, 2026: Quectel announces the SP900BF-AP smart module.
Engineering samples are made available to selected customers.
The Tech Race
The SP900BF-AP extends the Qualcomm Dragonwing IQ-9075 into a modular format that lets industrial device makers add edge AI without designing custom compute boards. It positions Quectel among module suppliers competing to make on-device AI practical for robotics and industrial systems.
Device makers building robots and industrial equipment could gain on-device AI without sending sensor data to the cloud, which can lower bandwidth costs. Pricing and availability for general purchase have not been announced, so near-term impact is limited to engineering evaluation.
The takeaway
Edge AI modules that process data locally are moving from prototypes toward industrial deployment. Teams evaluating robotics or industrial vision systems should weigh local processing against cloud costs and privacy requirements.
Further reading
For more on the chips powering devices like this, explore Semiconductors coverage.



