Google Partners With Jeff Bezos’ Blue Origin to Launch AI Data Centres Into Space by 2029

Share Us

155
Google Partners With Jeff Bezos’ Blue Origin to Launch AI Data Centres Into Space by 2029
20 Nov 2026
min read

News Synopsis

Google has partnered with Jeff Bezos’ Blue Origin to launch AI computing satellites in 2029, advancing Project Suncatcher’s ambition to develop solar-powered data centres in space.

Google and Blue Origin Join Hands for Space-Based AI Data Centres

Google is preparing to take its artificial intelligence infrastructure beyond Earth through a partnership with Jeff Bezos’ space company, Blue Origin. The two companies have announced plans to use Blue Origin’s New Glenn rocket to launch Google’s AI computing satellites into orbit in the first quarter of 2029.

The mission will support Google’s ambitious Project Suncatcher, a long-term initiative exploring whether a network of solar-powered satellites can perform AI computing operations in space. The satellites will carry Google’s Tensor Processing Units (TPUs), specialised chips designed to handle demanding artificial intelligence workloads.

Under the agreement, Blue Origin’s New Glenn rocket will transport multiple Google satellites into a sun-synchronous orbit. This type of orbit allows satellites to receive sunlight for extended periods, making it suitable for testing solar-powered computing systems.

The companies have not revealed the financial details of their partnership. However, the planned launch is expected to be an important milestone in Google's efforts to develop an alternative approach to building large-scale AI infrastructure.

What Is Google’s Project Suncatcher?

Project Suncatcher is Google's experimental programme to investigate whether artificial intelligence data centres can operate using computing satellites in space.

The central idea is to connect several satellites equipped with TPUs so that they can work together on machine-learning tasks. Each satellite is expected to offer computing power comparable to a rack of servers inside a traditional data centre.

Unlike ground-based facilities, these satellites could access solar energy for much of their time in orbit. Google believes that this approach could eventually help meet the growing electricity requirements of AI systems without depending entirely on Earth's power infrastructure.

Travis Beals, senior director of Project Suncatcher at Google, described the initiative as a long-term effort to expand AI infrastructure in space by using the abundant solar energy available in orbit.

The partnership with Blue Origin will support the next stage of the experiment by providing a dedicated rocket launch for Google's computing satellites.

Google Has Already Tested Its First AI Satellite

Google has already taken an initial step towards its space-based computing ambitions. In October 2026, the company launched its first Project Suncatcher prototype aboard a SpaceX Falcon 9 rocket.

Unlike the upcoming Blue Origin mission, the earlier launch was part of a rideshare operation that carried approximately 130 payloads into space. Google's satellite was among the various payloads included in that mission.

Following the launch, Google conducted an experiment to demonstrate that its satellite could process AI requests while operating in orbit.

The company sent a prompt from Earth to its Gemma 4 artificial intelligence model running on the satellite. The model processed the request and returned an AI-generated haiku to Earth.

This experiment demonstrated that a satellite equipped with Google's TPUs could run an AI model in space and communicate its response back to the ground.

The prototype is also helping engineers understand how computing hardware performs under the demanding conditions of space. These tests will provide valuable information before Google attempts to deploy a larger network of interconnected satellites.

Why Does Google Want to Build AI Data Centres in Space?

The rapid expansion of artificial intelligence has increased demand for computing infrastructure around the world. Modern AI models require powerful processors, large data centres and enormous amounts of electricity to operate.

Building and expanding traditional data centres also requires significant land, reliable power supplies and advanced cooling infrastructure. These requirements could become increasingly difficult to manage as AI adoption grows.

Google believes that moving some computing operations into space could help address these challenges.

1. Access to Abundant Solar Energy

One of the biggest potential advantages of orbital data centres is access to sunlight. Satellites placed in suitable orbits could receive solar energy for extended periods, reducing their dependence on conventional electricity sources.

2. Reduced Pressure on Earth's Infrastructure

Large data centres consume substantial electricity and require extensive cooling systems. If some AI computing workloads can be shifted into orbit, it could potentially reduce pressure on electricity grids, land availability and other resources on Earth.

3. Connected Satellites for AI Computing

Google's long-term vision involves creating a network of satellites that can communicate and work together. Instead of relying on a single spacecraft, multiple satellites could share computing tasks and collectively support machine-learning operations.

However, these benefits remain experimental. Building a reliable and commercially viable orbital computing network will require further technological development.

Major Technical Challenges Facing Space-Based Data Centres

Despite the potential advantages, operating AI infrastructure in space presents several engineering challenges. Google's prototype mission is designed partly to understand whether specialised computing hardware can survive and function reliably in orbit.

Cooling Powerful AI Chips

Heat management is one of the biggest obstacles. Powerful AI processors generate considerable heat during operation. On Earth, data centres use fans, air-conditioning systems and liquid-cooling technologies to manage temperatures.

Space does not provide the same airflow-based cooling conditions. Engineers must develop systems that can transfer heat through specialised pipes and radiators or panels to prevent equipment from overheating.

Radiation and Hardware Reliability

Satellites also face exposure to radiation in space. High-energy particles can interfere with electronic components and damage computing hardware.

Radiation may also cause data corruption through a phenomenon known as a bit flip, in which a binary value changes unexpectedly. Such errors could affect the accuracy and reliability of AI calculations.

Launch Costs and Long-Term Investment

Sending satellites into orbit is expensive. Developing a large network would require multiple launches, specialised hardware and significant investment.

Although the New Glenn partnership provides Google with a launch opportunity, expanding the project from a small prototype to a large-scale computing network could take years.

Other Companies Are Exploring Orbital AI Data Centres

Google is not the only company interested in moving computing infrastructure into space. Several technology and space companies are exploring similar possibilities as demand for AI processing continues to increase.

Indian space startup Pixxel and AI company Sarvam AI are also working together on plans related to space-based data centres. Their collaboration reflects growing interest in combining satellite technology with artificial intelligence.

SpaceX CEO Elon Musk has also strongly supported the idea of orbital AI infrastructure. He has argued that electricity demand from artificial intelligence could eventually exceed what Earth-based infrastructure can comfortably support.

According to Musk, space-based computing could provide access to additional energy and computing capacity without placing excessive pressure on communities and the environment. However, not everyone agrees with this approach.

Why Some Industry Leaders Question Space-Based Data Centres

SoftBank CEO Masayoshi Son has expressed doubts about the commercial feasibility of building AI data centres in orbit.

His concern is that savings from solar energy and potentially lower electricity costs may not compensate for the overall expense of deploying and maintaining computing hardware in space.

Launching satellites, replacing damaged equipment and establishing a large interconnected network would require substantial financial resources.

For example, a SpaceX Falcon 9 launch costs approximately $74 million, according to the information provided in the report. Multiple launches would be necessary to establish a large orbital infrastructure network.

This raises an important question: will the long-term benefits of space-based AI computing justify its high initial investment and operational complexity?

Conclusion: Google’s Space AI Ambitions Enter a New Phase

Google's partnership with Blue Origin marks another significant step towards exploring AI computing beyond Earth. The planned 2029 New Glenn mission will help the company test whether solar-powered satellites equipped with TPUs can support demanding AI workloads.

However, challenges involving cooling, radiation protection, launch expenses and satellite networking remain unresolved. While orbital data centres could eventually provide an alternative source of computing capacity, their commercial viability will depend on successful testing and substantial technological progress.