The platform is expected to be used in a wide range of products, including robots, drones, industrial handheld devices, connected cameras, AI edge boxes, and intelligent displays.
Qualcomm has launched the Qualcomm® RB3 Gen 2 Development Kit, targeting developers aiming to build high-performance, AI-capable IoT devices. The kit caters to a variety of IoT applications such as industrial handhelds, tablets, human-machine interface systems, point-of-sale systems, drones, kiosks, edge AI boxes, and connected cameras. The Platform is available in two versions: the Vision Kit, which includes camera support up to 192 MP, and the Core Kit, which includes the main board and other accessories. The development kit is powered by the Qualcomm QCS6490 SoC, designed specifically for high-tier IoT solutions.
The key features of the kit are:
- Octa-core Qualcomm Kryo 670 CPU
- Adreno 643 GPU
- Qualcomm Hexagon processor supporting 12 TOPS
- Enables a 10x increase in AI processing capabilities compared to the previous generation
- Supports on-device machine learning combined with edge computing
- Connectivity options: Wi-Fi 6E, Bluetooth 5.2 and LE audio
The kit supports Qualcomm Linux, which includes OS, software, tools, and documentation designed for Qualcomm IoT platforms. Developers benefit from direct support and full control over the operating system, allowing for maintenance and upgrades. This flexibility is not available with downstream versions of the Linux Kernel.The Platform enables developers to integrate pre-optimized AI models into their applications, reducing time-to-market and unlocking benefits such as immediacy, reliability, privacy, personalization, and cost savings. It is now available for pre-order in two integrated development kits and supports downloadable software updates to simplify application development, integration, and prototyping.
It is supported by the recently announced AI Hub, which contains a library of pre-optimized AI models for superior on-device AI performance, lower memory utilization, and power-optimized operation. Developers can view a selection of models and integrate them into their applications, enhancing the benefits of on-device AI implementations. The platform features support for Linux, providing a unified Linux distribution for multiple SoCs. The Qualcomm Linux software stack supports all processor cores, subsystems, and components within the platform.
