Himax Technologies, Inc. announced that the Company, in collaboration with Seeed Studio, the IoT hardware partner providing services that empower developers to implement their projects and products for digital transformation, will unveil a cutting-edge battery-powered endpoint AI vision processing module, Grove Vision AI Module V2, at CES 2024. This processing module features Himax's WiseEye2 AI processor HX6538 ("WE2"), renowned for its extremely low power consumption and exceptional AI inference performance, offering the most versatile and extensive AI development capabilities. Positioned at the forefront of the industry, the processing module serves as a powerful endpoint AI development platform, paving the way for future AI applications.

The new generation WE2 AI processor is built upon the success of its predecessor chip, WE1. WE2 leverages the advanced Cortex-M55 and Ethos U55 architecture to deliver 32 times faster inference speed and 50 times greater energy efficiency when compared to WE1. These enhancements allow the WE2 to enable more endpoint AI computational requirements while maintaining ultralow power consumption.

The WE2 features a multi-layered power management structure and incorporates Dynamic Voltage Frequency Scaling (DVFS) technology, to achieve microampere level power consumption. WE2 also incorporates impressive security features, including a Physical Unclonable Function (PUF) security mechanism that ensures each chip possesses a unique identity, mitigating the risks of unauthorized access and counterfeiting, thus enhancing the chip security. Additionally, it offers universal interface design that supports numerous IO interfaces, allowing developers to connect different types of external devices on the same interface, and greatly enhancing the flexibility and convenience of product development.

For software development, Seeed Studio's Endpoint AI vision processing module not only provides dozens of pre-trained AI models but also features the EdgeLab AI toolchain. This powerful tool chain allows users to tailor AI model trainings according to their specific requirements, thereby achieving more personalized AI software developments. When it comes to system-level hardware and software Interoperability and collaboration, the Seeed Studio processing module can seamlessly integrate with the widely popular Seeed's universal XIAO microcontroller development board series.

The combination enables the integration of the processing module into various household appliances with No-Code AI methodology, making the deployment of AI technology in home appliances more efficient.