Chinese semiconductor manufacturer Hygon Information Technology has introduced a new line of central processing units (CPUs) tailored for industrial robotics and automation devices, marking the company’s expansion from cloud computing into the growing field of physical artificial intelligence (AI).

Unveiled on Tuesday in Shenzhen, the Hygon 1000 series processors are designed for integration into industrial automation systems such as robotics and factory controllers. These CPUs enable local computing and graphics processing, reducing reliance on remote data centers. The chips feature four cores and eight threads, are compatible with both DDR4 and DDR5 memory, and include integrated graphics capabilities—Hygon’s first processors to do so. The new chips are notably compact, with package sizes nearly half that of the 3450M processor launched last year for laptops, and are priced between US$40 and US$80.

“At this moment, computing power is clearly shifting from cloud-based centers to edge and device-level computing,” said Sha Chaoqun, CEO of Hygon, at the launch event. He emphasized that AI applications are increasingly moving out of data centers and into physical environments such as factories, power stations, rail systems, and medical devices.

Hygon is positioning the 1000 series to meet the stringent demands of edge computing, where processors must deliver real-time performance, maintain low power consumption, and uphold security in diverse and sometimes unpredictable conditions. Huang Wei, general manager of Hygon’s Innovative Product Department, noted the rapid growth of the edge AI market and identified the industrial control sector as a primary target for the new embedded processors. He highlighted the role the chips will play in the expanding integration of AI within industrial robotics and smart manufacturing operations.

The new processors broaden Hygon’s existing CPU lineup, which includes the high-performance 7000 series geared toward data centers, the mid-tier 5000 series for industry cloud and edge applications, and the cost-efficient 3000 series. The company also produces deep computing units, accelerator cards designed specifically to enhance AI workloads. Huang indicated plans to launch the next generation of the 1000 series in 2027, with an anticipated doubling of computing performance.

Hygon’s move into embedded edge-computing chips comes amid ongoing efforts by China to achieve semiconductor self-sufficiency, particularly in the face of U.S. restrictions limiting China's access to advanced foreign chip technology. The Shanghai-listed company reported a 66.5 percent increase in first-half revenue compared to the previous year, reaching 9.1 billion yuan (approximately HK$10.6 billion).

This latest development signals Hygon’s strategic focus on expanding beyond data centers into AI-enabled physical devices, aligning with broader technological trends in industry and infrastructure.