Bestom is a Rockchip AIoT IDH design partner. We don't just sell core boards — we deliver Edge AI Host + Thin Clients + Cloud Collaboration, a complete AI transformation package that fits any small-commercial or industrial scenario. Built for engineers, PMs, and AI teams who need references they can cite.
We work alongside hardware teams in the regions we know best — from Shenzhen's consumer-electronics labs and Taipei's EMS clusters, to Tokyo's R&D centers, Seoul's robotics startups, Munich's industrial automation, and the Bay Area's edge-AI pioneers. Local time zones, local references, same citable engineering.



From 6 TOPS flagship edge AI to battery-free sensor nodes, on a unified BSP and RKNN deployment path.
Edge AI servers, on-device large models (7B/13B), robotics, smart cockpit, industrial HMI.
NVR, AI box, industrial HMI, AI POS, lightweight robotics.
Mid-high AIoT, commercial display, robotics main controller.
Industrial gateway, HMI, NVR, commercial display, self-service terminal.
Tablet, e-reader, smart panel, lightweight terminal.
Cost-sensitive entry-level AIoT devices.
Every solution ships with downloadable reference designs and SDK guides.
RK3576/3588 + RK1828 carries the large model; HMIs, sensors, and controllers become thin clients. Deployable to any small-commercial or industrial scenario.
RV1126 / RV1106AI-ISP full-color low-light, object detection, face recognition. IPC, access control, and behavior analytics reference designs.
RK3568 / RK3506Wide-temperature industrial grade, multi-serial (CAN/UART), dual Ethernet. HMI and protocol gateway in mass production.
Microphone array, offline voice, audio DSP. M08D module (RK2108D + RK962) for smart screens and soundbars.
RK3588M / EN50155Rail-vehicle terminals F901/902/903 in long-term supply, EN50155 vibration and wide-temperature compliant.
Battery-free nodesProprietary Ant protocol + energy harvesting. Battery-free wireless sensor network for asset and environment monitoring.
Each platform ships with Datasheet, schematic reference, BSP/driver, and RKNN deployment guide — usable directly in your design review and procurement docs.
From RK3588 to RV1106, NPU inference shares the RKNN-Toolkit2 quantization flow. Train once, deploy across the lineup.
Industrial-grade −40~85°C, EN50155 rail experience, CE/FCC compliance docs ready. Designed for long product life cycles.
SoM → eval board → small batch → mass production. Full IDH flow to shorten your time-to-market.
Send us your project brief: target scenario, target SoC, expected volume, and timeline. We'll respond within one business day.
Start a Project →