Infineon and Xiangshuo unveil 20 Gbps USB, setting AI camera edge standard

Infineon Technologies and ASMedia Technology announced a strategic partnership on the 27th of this month to jointly advance the application of USB 20Gbps high-speed interface technology in machine learning cameras and physical AI systems. The core of this collaboration lies in combining Infineon’s mature solutions in high-speed data transmission and power management with ASMedia’s technological strengths in AI inference chips and edge computing platforms. Together, the two companies aim to build a complete product chain characterized by low latency and high throughput, thereby meeting the growing demands of the AI imaging and sensing markets.

Infineon, a prominent German semiconductor supplier, has long focused on power semiconductors, automotive electronics, and industrial control chips. In recent years, the company has accelerated its entry into the AI sector by launching multiple AI accelerators and autonomous driving-related modules, while continuing to innovate in high-frequency interface technologies. As a next-generation high-speed interface standard, USB 20Gbps offers a data transfer rate of 20 gigabits per second, which is particularly critical for applications requiring real-time image processing and deep learning inference. Infineon possesses mature hardware design and verification experience in this technology, enabling it to provide partners with a reliable interface foundation.

ASMedia Technology specializes in the research and development of AI chips and edge computing platforms, with products widely applied in smart cameras, industrial robots, and automated inspection. The company excels at integrating high-performance AI algorithms with low-power hardware to achieve real-time inference and data processing. As the machine learning camera market expands rapidly, ASMedia’s AI chips have become core components in many high-end imaging systems. Through its partnership with Infineon, ASMedia can further enhance the data transmission performance of its chips and reduce the overall latency from sensing to inference.

On the technical front, the two parties will jointly develop high-speed data channels based on USB 20Gbps, closely coupling Infineon’s power management and signal integrity solutions with ASMedia’s AI inference chips. This approach not only improves overall system performance but also reduces power consumption and costs, which is particularly important for smart cameras and industrial robots that require high-performance AI inference within limited spaces. Over the next 12 to 18 months, the two companies plan to launch a series of integrated products and begin mass production for automotive safety cameras, industrial inspection cameras, and smart surveillance systems.

From a market perspective, global demand for AI imaging and edge computing continues to rise. According to reports from market research firms, the machine learning camera market size is expected to exceed $20 billion in 2025, with an annual growth rate of 15%. Furthermore, with the advancement of autonomous driving and smart manufacturing, the need for low-latency, high-throughput data transmission standards is becoming increasingly urgent. The partnership between Infineon and ASMedia aligns perfectly with this trend, allowing both companies to secure a larger share of the international market while providing new revenue opportunities for the Taiwan semiconductor industry.

This collaboration also holds significant importance for the Taiwan semiconductor ecosystem. Taiwan plays a key role in the global chip supply chain, particularly holding a leading position in fabrication and packaging technologies. Through its cooperation with Infineon, ASMedia can further enhance the international competitiveness of its products while injecting more high-value-added components into the local supply chain. In the long run, the synergistic effects between the two parties will promote innovation and industrial upgrading in Taiwan within the fields of AI chips and high-speed interface technology, positioning the country more favorably in the global AI ecosystem. (Source: Central News Agency)