GigaDevice has announced it will showcase advancements in AI-driven edge applications and digital power solutions—targeting data center and server environments—at Embedded World North America from November 4 in Anaheim, California. The demonstrations highlight technical capabilities of its latest 32-bit microcontrollers (MCUs) across facial and voice recognition, high-efficiency power management for AI servers, and compact interactive consumer devices.
For data center and server operators, GigaDevice will demonstrate a 12 kW artificial intelligence server power system, developed in collaboration with Navitas Semiconductor via their Digital Power Joint Lab. The design uses GigaDevice’s GD32G553 MCU alongside Navitas GaNSafe integrated circuits and third-generation Fast silicon carbide (SiC) MOSFETs. According to GigaDevice, the system supports both Open Compute Project (OCP), Open Rack Version 3 (ORv3), and Common Redundant Power Supply (CRPS) standards, exceeds the 80 PLUS “Ruby” efficiency level, and achieves a 97.8 percent peak efficiency. Target applications include hyperscale data centers, AI servers, and traditional server racks.
GigaDevice will also present edge AI recognition systems using its GD32H7 MCU series. Facial recognition is executed on the open-source Swift Yolo model, which achieves a mean average precision of 97.9 percent and a single inference time of 77 milliseconds on the GD32H759 development board. Its voice recognition system integrates the Sensory engine for multi-keyword and multilingual support.
In consumer electronics demonstrations, GigaDevice will show a feature-rich system for compact devices centered on the GD32C231 MCU, equipped with a 1.47-inch TFT-LCD, touch control, haptic feedback, and advanced user interface capabilities.
GigaDevice reports that it has formed strategic alliances with foundries and assembly/test facilities to streamline the supply chain and holds certifications including ISO26262:2018 ASIL D for automotive safety and IEC 61508 for functional safety.
Source: GigaDevice
