At the "Xingsuan Zhilian" Space Computing Power Seminar in 2026, Zhang Ping, a counselor of the State Council and a professor at Beijing University of Posts and Telecommunications, delivered a keynote speech, deeply analyzing the core logic of the coordinated development of satellite internet, space computing power, and 6G. He proposed that "space computing power + 6G" will become the core path to disrupt the future, and elaborated on the development direction and practical achievements of the integration of communication and AI (ComAI), providing new dimensions of thinking for the global technological competition landscape.

As a key technology that breaks through the physical limitations of traditional terrestrial networks, satellite internet can achieve high-speed, low-latency, and highly reliable internet access globally, especially covering remote areas, oceans, high altitudes, and other regions that are difficult to reach by terrestrial networks. With the iteration of aerospace technology and the large-scale launch of commercial satellites, the civilian value of satellite internet continues to be highlighted. From the early Iridium satellite program, to the 03b constellation, Starlink system, and then to the domestic "GW" constellation and Qianfan constellation, the industry has undergone multiple rounds of construction boom and has become a strategic field that major countries in the world are competing to develop.

Entering the AI era, the capability boundaries and development space of the Internet plus have been further expanded. Relying on high- and low-orbit satellites and 5G/6G network technologies, the Internet plus has achieved breakthroughs in low-latency and high-speed data transmission; while the deep empowerment of AI technology has optimized data transmission and management efficiency, significantly enhancing network adaptability and security. It is noteworthy that the Internet plus is a product of multidisciplinary integration. Inspired by the natural intelligence information processing flow, the deep integration of communication and intelligence (ComAI) is becoming a mainstream trend in the industry. Its core competitiveness lies in the distribution, update, and evolution of semantic carrying models within the network. Through a three-tier architecture of "semantic information aggregation - large model layer processing - on-demand update", a more intelligent and adaptable network ecosystem is constructed.

However, the shortage of computing power has become the core bottleneck restricting the development of ComAI. Traditional terrestrial computing power not only faces the practical problem of excessive power consumption but also struggles to meet the endless demand for computing power in space scenarios. In response, Zhang Ping pointed out that the key to breaking this deadlock lies in the development of space-based computing power and distributed computing. The collaborative linkage between space computing power and 6G is the core path to solve this dilemma and disrupt traditional communication models. This collaborative model will bring multiple changes: low-orbit satellites will achieve global seamless networking, solving the problem of poor terrestrial communication coverage; space computing power will provide fast and low-cost computing support to meet the rigid demand for computing power in space scenarios; multi-modal sensing-communication-guidance and computing will be deeply integrated, enabling the "module + computing" terrestrial terminal mode to be implemented in space scenarios; space-based computing power, combined with satellite-to-ground and inter-satellite laser technology, will weave a space-based computing power network with full coverage, promoting a new development pattern of "supercomputing in space", "cloud computing in space", and "services in space".

At the technical implementation level, semantic communication is clearly identified as the fundamental approach to achieve ComAI. Unlike traditional communication, which pursues lossless and efficient transmission, semantic communication focuses on higher-level semantic information transfer, and its channel capacity can exceed the theoretical limits of Shannon's information theory. Zhang Ping highlighted four key supporting technologies: First, joint source-channel coding, which utilizes processes such as semantic extraction and JSCC encoding and decoding, combined with video context loop enhancement to significantly improve the efficiency of satellite-to-ground video transmission. Second, semantic knowledge base, which achieves semantic compression based on a shared codebook and ensures transmission robustness through deep neural networks, effectively overcoming the "cliff effect" under low signal-to-noise ratios. Third, semantic-level feedback retransmission, which uses semantic mutual information as a feedback indicator to achieve selective transmission and adaptive retransmission, optimizing system robustness and transmission efficiency. Fourth, high-robustness transmission, which balances computational resources and transmission efficiency through three layers of modules: semantic importance segmentation, robustness protection, and adaptive transmission, reducing bandwidth occupation by over 30%.

Multiple practical verifications have fully demonstrated the technical feasibility. In May 2025, Beijing University of Posts and Telecommunications (BUPT) collaborated with China Telecom to conduct a full-chain self-developed video semantic communication satellite verification. It achieved efficient transmission of 720P standard definition video in narrowband and low signal-to-noise ratio (SNR) scenarios, with a transmission efficiency increase of 3 times, and achieved a 7DB gain in low SNR scenarios. Furthermore, BUPT has established a long-term strategic cooperation with Guoxing Aerospace, relying on the space computing power platform of the "XingSuan" plan constellation to build a semantic native ubiquitous intelligent and simple wireless air interface link, promoting the construction of semantic satellite constellations. It is expected to achieve significant results in terms of communication efficiency improvement by 50%, business service quality enhancement by 20%, resource consumption reduction by 30%, and latency reduction by 20%.

Looking ahead, Zhang Ping emphasized that collaborative efforts from industry, academia, research, and application are the core path for technology implementation. With the continuous breakthroughs in space computing power and 6G collaborative technology, as well as the large-scale application of key technologies such as semantic communication, it will drive disruptive changes in the communication industry and inject new momentum into global technological development.