China's Orbital Chenguang Secures $8.4 Billion in Credit Lines for Space Data Centers

A Multibillion Dollar Bet on Computing in Orbit

Beijing Orbital Twilight Technology Co., Ltd., which operates as Orbital Chenguang, has secured access to 57.7 billion yuan, approximately $8.4 billion, in strategic credit lines for an ambitious project: building data centers in space. Twelve major Chinese banks, including Bank of China, Agricultural Bank of China and CITIC Bank, are backing the financing capacity.

The late April announcement places one of the largest financial commitments yet to infrastructure in space behind a startup seeking to move part of the world's computing workload off Earth. Its target is an orbital data center with power capacity exceeding one gigawatt by 2035, using satellites flying between 700 and 800 kilometers above the planet.

A credit line is money a borrower can potentially draw upon as needed. The 57.7 billion yuan therefore represents access to financing, rather than cash already transferred to the company. Even with that qualification, the scale of the commitment and the participation of major lenders show substantial institutional support.

Orbital Chenguang is pursuing a response to pressures already facing data centers on Earth. These facilities power social media, artificial intelligence training and other digital services, but they require land and large amounts of electricity while producing considerable heat. More complex artificial intelligence models are increasing demand for computing capacity.

The company's proposal is to place computing equipment where solar energy is abundant and physical space is less constrained. Its plan combines a startup, public research institutions, a consortium of industrial partners and bank financing. Turning that arrangement into a working commercial service will require progress in power generation, heat disposal, communications and launches.

A Small Equity Round Beside a Large Credit Facility

Orbital Chenguang completed its Pre A1 funding round on April 20, attracting venture capital and industrial investors including Haisong Capital, Cathay Capital and InnoAngel Fund. A funding round of this kind brings investment into a company in exchange for ownership stakes.

The equity investment amount has not been disclosed. Supplementary reporting suggests it was below $1 million, a modest sum beside the approximately $8.4 billion in potential bank financing.

The contrast explains the company's funding structure. Immediate private investment is much smaller than the debt capacity available through the banks. Drawing on that capacity would create borrowing obligations, making the project's eventual ability to generate income central to its commercial prospects.

Orbital Chenguang was incubated by the Beijing Astro future Institute of Space Technology. The institute has government backing through support from Beijing's municipal science and technology commission and the Zhongguancun Science Park administration.

Zhang Shancong is both the institute's director and Orbital Chenguang's chief scientist. His two roles connect public research work with the commercial company tasked with putting the technology into operation.

The institute also leads a consortium of 24 organizations covering the industrial chain, from component manufacturing to launch services. That structure gives the initiative a broader base than a startup working alone: spacecraft components, computing hardware and access to orbit must all be coordinated for the proposed system to function.

Why the Plan Uses a Dawn Dusk Orbit

Orbital Chenguang's blueprint calls for a constellation, or coordinated group of satellites, operating in a dawn dusk orbit at an altitude of 700 to 800 kilometers. This is a particular form of Sun synchronous orbit, in which satellites maintain a consistent orientation relative to the Sun.

The choice is intended to provide near continuous sunlight for solar arrays while avoiding the extreme temperature changes encountered in other orbital paths. Solar equipment above the atmosphere can collect energy without weather interruptions or atmospheric filtering.

Zhang Shancong, director of the Beijing Astro future Institute of Space Technology and Orbital Chenguang's chief scientist, described the pressures on facilities on Earth during a November briefing. He identified land, electricity and cooling as constraints on further expansion:

Large scale data centers have expanded rapidly worldwide, but further growth faces major obstacles, including heavy land use, soaring energy consumption and limits on atmospheric cooling.

Orbital computing would seek to bypass those constraints by moving equipment and power generation into space. The company envisages solar arrays supplying electricity and passive thermal management helping the equipment release heat.

The 2035 target exceeds one gigawatt of power capacity. A single gigawatt is comparable to the output of a large nuclear reactor. Providing that much power in orbit would mean building an industrial computing complex, rather than simply testing a few computers aboard satellites.

The constellation could potentially require thousands of spacecraft, depending on how much power each satellite generates. The size of the fleet would therefore depend on the capabilities of individual spacecraft as well as the total computing capacity sought.

National Planning and the Schedule to 2035

Orbital Chenguang's ambitions align with China's 15th Five Year Plan, the national economic blueprint guiding industrial development through 2030. The project sits within a wider effort to connect commercial space activity with national technology priorities.

In January, China Aerospace Science and Technology Corporation, or CASC, proposed developing computing infrastructure in space at gigawatt scale. CASC is China's principal space contractor, and its proposal included an integrated cloud edge terminal architecture in orbit.

That architecture connects large central computing resources, processing closer to where data is collected, and the devices that send or receive information. In practical terms, it would distribute computing tasks across different parts of a connected system rather than send every task to a single central facility.

CASC's involvement suggests a possible coordinating role across commercial companies and organizations connected to the state, rather than responsibility for one unified program. Orbital Chenguang's institute and industrial consortium fit that broader approach.

The proposed development schedule stretches from initial technology work to a large operational facility:

• 2025 to 2027: address core technological barriers and deploy an initial computing constellation.
• 2028 to 2030: integrate data processing on Earth with computing resources in space.
• 2035: operate a fully functional orbital data center at large scale, with power capacity exceeding one gigawatt.

An Experimental Launch and Rocket Setbacks

Chenguang 1, an experimental satellite, was initially scheduled to launch in late 2025 or early 2026. Its status remains unclear as of April 23, 2026, with no confirmed launch reported.

The planned experiment forms part of the early phase of development, when the initiative must move from design work toward a functioning constellation. The schedule places this work within the 2025 to 2027 period for resolving core technical problems and beginning deployments.

China's commercial launch sector has also experienced setbacks this year. Satellites were lost during the debut flights of the Ceres 2 and Tianlong 3 rockets, adding pressure to plans that depend on access to space.

A constellation potentially numbering in the thousands would require repeated launches as well as spacecraft production. Rocket availability and reliability are therefore practical concerns throughout the development schedule, not simply at the moment of the first experimental mission.

Other Chinese Companies Are Already Testing the Idea

Orbital Chenguang is one of several Chinese ventures pursuing computing in orbit. Some are testing smaller systems or processing data aboard individual satellites, while others are developing more powerful spacecraft.

In May 2025, ADA Space and Zhejiang Lab launched 12 satellites for the Three Body constellation. The project takes its name from the famous science fiction series and focuses on edge computing, which means processing information closer to where it is generated.

For satellites, that approach can place computing equipment alongside the instruments collecting data. It is a different stage of development from Orbital Chenguang's proposed large data center, but it belongs to the same effort to carry out more processing in space.

Shanghai Bailing Aerospace Technology Co. has secured tens of millions of yuan in early funding. It plans to launch a demonstration satellite in 2026 and is developing computing satellites in the 100 kilowatt class.

Zhongke Tiansuan, also known as Comospace, has already deployed its Aurora 1000 computing technology aboard a Jilin 1 satellite. These projects show that the domestic field includes both future constellation plans and computing technology already placed in orbit.

International Ambitions and Vast Constellation Filings

Google has explored concepts such as Suncatcher for solar energy and computing applications in space. SpaceX has reportedly envisioned constellation designs that could eventually include one million satellites, although that figure remains an aspiration.

The international activity reflects growing interest in orbital infrastructure as a possible component of future computing capacity. The proposals vary greatly in size and maturity, from equipment aboard existing satellites to fleets far larger than those contemplated for early demonstrations.

China's ambitions also extend to securing resources for very large satellite networks. In December, authorities filed documentation with the International Telecommunication Union for two proposed constellations, each covering 96,714 satellites.

No direct connection has been identified between those filings and Orbital Chenguang's data center initiative. The applications nevertheless show planning for massive future deployments and efforts to secure orbital positions and spectrum resources, the radio frequencies satellites need to communicate.

The scale of those proposals puts the computing project in a wider setting. Space infrastructure requires coordination of spacecraft, launches, communications and power, and several Chinese and international organizations are now considering how large such systems might become.

Why Cold Space Still Creates a Cooling Problem

Heat disposal is among the most demanding engineering problems facing orbital data centers. Computers turn much of the electricity they consume into heat, which must be removed to keep their components working.

Space offers a cold environment, but its vacuum changes the way cooling works. On Earth, air can carry heat away through convection. In orbit, there is no surrounding air to perform that job, so equipment must release heat through radiators.

Passive thermal management therefore still requires carefully designed hardware. At gigawatt scale, the waste heat from computing would demand massive radiator arrays, increasing the weight and complexity of the system.

The promise of abundant sunlight and the difficulty of cooling are connected. More electrical power allows more computing, but also creates a larger thermal load to manage. Solar arrays and radiators would both be essential parts of the proposed industrial installation.

Data transmission creates another constraint. Useful computing services would need radio frequency or optical connections capable of moving enough information between orbit and Earth to justify the expense of the system.

Laser communication technology is advancing, but remains expensive and sensitive to weather. Delays as signals travel to orbit and back would be manageable for many applications, while excluding tasks that need responses within milliseconds.

Orbital computing would therefore have to match its services to its communications capabilities. Processing power alone would not determine whether users could move their workloads into space at a useful speed and cost.

Launch Costs, Replacement Cycles and Competition on Earth

Launch costs are falling, but sending thousands of satellites into orbit would still be expensive. A working constellation would also require continuing manufacturing and launches after its initial deployment.

Satellites in low Earth orbit experience orbital decay and require replacement every five to seven years. For a large fleet, that creates a recurring burden rather than a single construction bill.

Most orbital equipment cannot be maintained once deployed. Systems would need to operate autonomously, with failures potentially permanent. Reliability would affect both the computing service and the number of replacement spacecraft needed.

Data centers on Earth are also improving. Better cooling technologies, greater use of renewable energy and more efficient chip designs may reduce some of the pressures that orbital facilities are intended to address.

Critics point to those improvements as a commercial challenge: the advantages of moving computing into space may narrow before a large orbital facility is ready. Orbital Chenguang's economics will depend on how its costs and performance compare with facilities on Earth by the time it reaches operation.

The bank financing makes those questions especially consequential. Access to 57.7 billion yuan could support an extensive development effort, while the resulting infrastructure would still need a workable business model to sustain borrowing, replacements and operations.

A Test of China's Approach to Strategic Technology

Orbital Chenguang's financing signals China's determination to build a position in the next generation of computing infrastructure. Artificial intelligence capabilities are increasingly important to economic and military competition, making the supply of computing power a strategic concern.

The project brings commercial execution together with a research institute supported by public authorities, an industrial consortium and major bank lenders. Private innovation is being placed within a national planning framework.

This hybrid approach differs from the predominantly private model seen in Western space industries. It may help China assemble resources across institutions more quickly, while leaving questions about market efficiency and whether the project can support itself commercially rather than depend on subsidized strategic capacity.

The attempt will create demand for launches and produce information about the technical limits of computing in orbit. A successful facility could become a model for further industrial activity in space; an expensive development effort could still provide lessons about power, cooling and communications.

The immediate position is more concrete than the distant ambition: Orbital Chenguang has substantial potential financing, a research and industrial network, and a schedule ending in 2035. Delivering the proposed facility now depends on converting those resources into reliable spacecraft, sufficient launch capacity and computing services that customers can use economically.

Key Points

• Orbital Chenguang has secured 57.7 billion yuan, approximately $8.4 billion, in credit lines from 12 major Chinese banks.
• The company completed its Pre A1 funding round on April 20; the equity amount is undisclosed, with reporting suggesting less than $1 million.
• The project aims for power capacity exceeding one gigawatt by 2035, using satellites in dawn dusk orbit at 700 to 800 kilometers.
• The Beijing Astro future Institute of Space Technology supports the venture and leads a consortium of 24 organizations.
• The schedule covers initial technology and deployments in 2025 to 2027, integration with computing on Earth in 2028 to 2030, and full operation in 2035.
• Chenguang 1 was planned for late 2025 or early 2026, but no confirmed launch has been reported.
• Other Chinese ventures include ADA Space, Zhejiang Lab, Shanghai Bailing Aerospace and Zhongke Tiansuan, also known as Comospace.
• Cooling, communications, launch costs, spacecraft replacement and competition from improving data centers on Earth remain central challenges.