The commercial space industry is now vying for a share of the computing power market.

At 12:03 on September 20, the Long March 1-1 carrier rocket successfully launched nine satellites, including the "Supercomputing-1" satellite, into their designated orbits.

This launch reveals a new trend in on-orbit satellites: computing power is being integrated into the core payload of satellites, and satellites are evolving from data acquisition terminals to computing nodes.

"Supercomputing-1" is equipped with optical remote sensing and AI computing payloads.

According to the project team, the satellite can process images and identify targets in orbit, and only transmit the selected high-value data back to the ground. In some scenarios, the data response time can be reduced from several hours to minutes.

The first satellite launched from Shenzhen, launched simultaneously with the rocket, integrates 5G non-terrestrial network base stations, core networks, onboard AI computing, and laser and microwave communication payloads onto a single satellite, verifying the on-orbit collaboration of communication, sensing, and computing.

A day ago, Zhuzhou Space's PIESAT-2 satellites 13 to 16 entered their designated orbits. These SAR satellites carry onboard intelligent processing and mission planning payloads, enabling them to directly interpret remote sensing data in orbit and then output application results to the ground.

Within two days, multiple satellites with on-orbit processing capabilities were launched into orbit. The verification of space computing power is moving from scientific research demonstration to commercial and engineering deployment, and it also brings new mission sources for commercial satellites and rocket launches.

Satellite data processing is currently the most clearly defined application of space computing power.

When a satellite passes a single ground station, the effective communication window is typically only about 10 minutes, and the raw images generated by a high-resolution remote sensing satellite in a single operation can reach hundreds of gigabytes.

Limited communication windows and ever-increasing data volumes have made satellite-to-ground transmission a bottleneck for the efficiency of remote sensing applications. Zhejiang Lab previously disclosed that, limited by ground station resources and bandwidth, approximately 90% of satellite data failed to be transmitted back to Earth, with the average transmission time for remote sensing data exceeding one hour.

In-orbit computation can first identify fire points, ships, flood extent, or surface changes, and then transmit the coordinates, extent, and analysis results back to Earth. This processing method can reduce the amount of data and shorten the time between satellite observation and ground decision-making.

Emergency rescue, marine patrol, natural resource management, and ecological monitoring all have high requirements for data timeliness and are where the value of minute-level response is most easily demonstrated. These scenarios will become the first markets where space computing power is applied.

Domestic research has expanded from single satellites to computing constellations.

The first batch of 12 satellites of the Three-Body Computing Constellation, launched in 2025, achieved a maximum computing power of 744 TOPS for a single satellite and an overall on-orbit computing power of 5 POPS. Subsequently, the first batch of satellites established inter-satellite laser links and deployed 11 AI models for on-orbit operation.

Inter-satellite networking expands the application boundaries of space computing power. Multiple satellites can share computing and storage resources, allocate tasks in orbit, and invoke algorithms according to different regions and scenarios. Model deployment, task scheduling, and data collaboration thus become new capabilities for computing satellites.

This change is also reshaping the demand structure of commercial spaceflight. Computing satellites require higher-performance processors, more stable power supplies, faster inter-satellite communication, and continuously updated software systems. This demand will propagate along the satellite platform, onboard chips, laser communication, operating systems, and data services throughout the industry chain.

Engineering capabilities determine the speed at which space computing power expands.

Vacuum environments rely primarily on radiation for heat dissipation, and high-density computing increases the thermal control pressure on satellites. Radiation-resistant chips, power systems, and high-speed inter-satellite communication also directly affect the performance and lifespan of computing payloads.

Launch capability determines the deployment cost and networking efficiency of computing nodes. Reusable rockets, mass production of satellites, and standardized deployment will become important foundations for scaling up space computing power.

Space computing power is adding a new layer of value to commercial spaceflight. Rockets are responsible for sending computing nodes into orbit, satellites complete data acquisition and processing, inter-satellite networks connect distributed computing power, and ground applications receive the final results.

This link extends the competition in commercial spaceflight from the number of satellites to data processing efficiency.

As more computing satellites enter orbit, participants who can more quickly form network capabilities, match high-time-efficiency scenarios, and continuously deliver data services will gain industrial initiative in space computing power.

Risk Warning and DisclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.