MediaTek, Raydium, Cloudminds Unveil Edge AI Chips for 50% Performance Boost

MediaTek and Raydium Semiconductor have recently introduced new chip products that are poised to significantly enhance edge AI capabilities. These advancements are part of a broader industry trend where companies are increasingly focusing on developing chips that can support AI applications at the edge, closer to where data is generated.

The new chips from MediaTek and Raydium Semiconductor are designed to handle complex AI tasks with greater efficiency and speed. This is achieved through a combination of advanced processing power and optimized algorithms, which allow the chips to perform real-time data analysis and decision-making. The new products are expected to be particularly useful in applications such as autonomous vehicles, smart cities, and industrial automation, where low latency and high reliability are critical.

MediaTek has unveiled its latest 3nm AI PC chip, the Kompanio Ultra, which boasts an impressive AI processing power of 50 TOPS. This chip is designed to enhance the generative AI experience on edge devices. The Kompanio Ultra features MediaTek's eighth-generation NPU and a full-core CPU architecture, including an Arm Cortex-X925 processor with a maximum frequency of 3.62GHz. This configuration ensures superior multitasking capabilities and efficient power usage. MediaTek has collaborated closely with Google to ensure that the latest Chromebook Plus models benefit from advanced edge AI capabilities, superior power efficiency, and immersive multimedia features.

In addition to the Kompanio Ultra, MediaTek has also introduced the MediaTek Pentonic 800, the world's first 4K 165HDR TV AI chip. This chip offers a 50% improvement in AI performance and a 60% reduction in memory bandwidth usage compared to its predecessor. The Pentonic 800 is equipped with MiraVision Pro intelligent display technology, providing an exceptional visual experience with a variable refresh rate of up to 165Hz. It supports various AI-enhanced image quality technologies, including AI Super Resolution (AI-SR 3.0), AI-Contrast 2.0, AI scene recognition, and AI object recognition. These features make the Pentonic 800 suitable for a wide range of applications, including smart TVs, projectors, displays, and embedded large screens.

Raydium Semiconductor has also made significant strides with its new solutions for the automotive industry. The company has showcased the RK3588M, an 8nm SoC integrated with edge-side large models, and the RK2118M/RK3308M, innovative automotive AI audio processing platforms. These solutions are designed to enhance local AI processing capabilities in smart vehicles, enabling advanced features such as multimodal interaction, scene perception, and personalized services. The RK3588M and RK3576M, combined with mainstream edge-side large models, offer a comprehensive solution for smart vehicles, supporting seamless switching between QNX and Android systems and improving resource scheduling efficiency.

Cloudminds has introduced its new DeepEdge 10 chip platform, designed for edge AI inference. This platform offers a maximum computing power of 48 TOPS and supports D2D/C2C Mash interconnection, making it capable of deploying large models with trillions of parameters. Cloudminds has developed several edge AI products based on this platform, including intelligent AI panoramic cameras and smart parking solutions. These solutions provide real-time parking occupancy detection, anomaly detection, and precise vehicle recognition, supporting scientific parking allocation. The DeepEdge 10 platform significantly reduces deployment costs by requiring fewer edge devices compared to similar solutions from other major manufacturers.

These advancements by MediaTek, Raydium Semiconductor, and Cloudminds highlight the growing importance of edge AI in various industries. As the demand for edge AI solutions continues to rise, these new chips are expected to drive innovation and enable a wide range of new applications, from smart homes and wearable devices to industrial robots and autonomous drones. The ongoing innovation in the semiconductor industry underscores the critical role of edge AI in shaping the future of technology.