Google books Blue Origin for 2029 orbital data center launch
Blue Origin’s New Glenn will launch Google’s Project Suncatcher satellites in 2029, testing solar-powered orbital AI compute as U.S. heavy-lift capacity fills through 2030.

Why it matters
- Blue Origin will launch Google’s Project Suncatcher satellites on a 7×2 New Glenn in 2029.
- New Glenn has launched three times to date; 7×2 is its smaller variant.
- U.S. commercial heavy-lift rockets, including New Glenn, are filling capacity through 2030.
- Each satellite targets roughly rack-scale computing in Sun-synchronous orbit with nearly continuous solar energy.
- Google did not name the satellite builder; the satellites may weigh several tons.
By Marcus Bennett
Blue Origin will launch Google’s Project Suncatcher satellites aboard a 7×2 New Glenn in 2029, Blue Origin said Thursday. The booking gives Google a dedicated launch for a planned test of solar-powered AI computing in orbit.
What will the 2029 mission demonstrate?
Google disclosed roughly a year ago that it was researching scalable data centers in orbit and described the effort as one of its research “moonshots.” The New Glenn mission is designed to deploy the satellites into Sun-synchronous orbit, where they can receive nearly continuous solar energy.
Blue Origin’s announcement says each spacecraft would carry computing capacity at roughly the scale of a rack in a terrestrial data center. That rack-scale comparison anchors Google’s proposed orbital hardware to a familiar unit of ground-based compute. The stated likely test would involve solar-powered satellites equipped with AI chips.
The likely objective for 2029 is to test whether those satellites can network with one another and demonstrate machine-learning compute, meaning AI processing, in space. The launch announcement frames the flight as a technology demonstration and provides no commercial service schedule.
That framing leaves the mission with a concrete technical purpose: establish whether multiple rack-scale, solar-powered systems can communicate and process AI workloads away from Earth. It does not by itself establish the cost of doing so.
Travis Beals, senior director of Project Suncatcher at Google, said: “Project Suncatcher is a long-term moonshot to scale AI infrastructure in space by tapping into the abundant solar power in orbit.”
He added: “This dedicated launch with Blue Origin represents a crucial milestone on our roadmap as we expand the boundaries of what’s possible for the future of compute.”
What remains undisclosed?
The announcement leaves the hardware and commercial arrangement largely undefined. Google did not identify the company contracted to build the satellites, and the spacecraft may weigh something on the order of several tons.
- Satellite count: The announcement does not state how many satellites the 2029 mission will carry.
- Onboard hardware: Google names AI chips as the computing component, but provides no chip supplier, benchmark, processing rate or power figure.
- System design: The announcement provides no deployment sequence or details for linking the satellites to ground systems.
- Price: Neither the launch price, satellite contract value nor projected cost per rack of orbital compute is disclosed.
- Schedule: 2029 is the stated launch year; the announcement does not narrow the date or set a deployment timetable for a larger constellation.
Those omissions do not make the 2029 flight trivial. They define its evidentiary limit: a successful network and machine-learning demonstration could validate the architecture while leaving the business case to later evidence.
Because Google has not named the builder or disclosed the satellites’ final mass, outside observers cannot judge from the announcement whether the design fits the mission or what the system will cost. The announced 2029 claim is narrower: demonstrate the core functions. The broader proposition concerns a possible constellation in the 2030s.
Why does the New Glenn booking matter?
Google’s mission will fly on the 7×2 variant of New Glenn, the smaller version of a rocket that has launched three times to date. The announcement provides no maximum payload mass or launch price for the 7×2 configuration.
Blue Origin’s announcement comes as U.S. commercial heavy-lift rockets, meaning large-capacity launch vehicles, including New Glenn, fill capacity through 2030. The 2029 slot therefore places Project Suncatcher inside a period of tightening launch availability.
Securing a named provider and rocket configuration gives Google schedule visibility for the test. It does not prove Blue Origin has reached the ultra-low launch costs required for a competitive orbital data center.
The booking also does not remove the need for repeat flights if Google pursues a large constellation in the 2030s. One dedicated launch is intended to demonstrate networking and compute; a large orbital deployment would require a sustained sequence of launches and the associated capacity.
Can space-based data centers beat terrestrial facilities?
The economic viability remains unresolved. The market assessment behind the Blue Origin-Google announcement calls this a point of broad agreement: without ultra-low-cost, heavy-lift capability, orbital data centers cannot compete economically with terrestrial data centers.
That condition links the 2029 engineering test to the larger question of a large orbital data-center constellation in the 2030s. Solar power can supply the systems; cheap access to orbit must supply the economics.
Several-ton satellites would raise the launch requirement further. New Glenn supplies a launch option for one test, but the announcement offers no evidence that a dedicated flight carries the ultra-low cost required for scaled deployment.
The solar premise comes directly from Beals: Google intends to tap “the abundant solar power in orbit.” Yet the same announcement supplies no launch price, no satellite contract value and no cost comparison with terrestrial racks.
Google faces a structural constraint because it does not have its own launch vehicle. If it pursues a large constellation, the company must secure future heavy-lift capacity from providers such as Blue Origin before it can turn a 2029 demonstration into a 2030s deployment.
That dependence makes launch availability and launch price central to Google’s long-term roadmap. It also connects the Blue Origin booking to a market whose commercial heavy-lift capacity is already filling through 2030.
What would count as a successful test?
Google’s likely technical test has two components. A third condition, economic competitiveness, will determine whether a successful demonstration becomes a viable business.
- Connectivity: Establish that solar-powered AI satellites can operate as a networked system in Sun-synchronous orbit.
- Compute: Demonstrate machine-learning processing in orbit at roughly rack-scale capacity per satellite.
- Commercial follow-up: Disclose enough launch, satellite and system costs to compare orbital computing with terrestrial data centers.
The announcement sets no numeric benchmark for either technical milestone and does not define how many networked satellites must operate together. A flight can therefore demonstrate feasibility without quantifying the performance or economics of a large constellation.
The 2029 flight sits between two different claims: a near-term systems test and a long-term infrastructure strategy. The first can be judged by connectivity and machine-learning operation in orbit. The second requires repeated launches, a satellite supplier and cost data that the current announcement does not provide.
That gap is the long-term problem behind the launch booking. The booking assigns Blue Origin the demonstration mission. Google still has to prove that the hardware can scale, that launches can remain available, and that orbital compute can match the economics of terrestrial computing.
The next concrete checkpoint is the 2029 New Glenn flight. Its success could show that Google’s rack-scale solar computing concept can connect and run AI workloads in orbit. It would not settle the commercial case unless Google and Blue Origin also disclose the launch and system costs that determine whether orbital computing can compete with terrestrial computing.
For Project Suncatcher to move from research moonshot to competitive infrastructure, the 2029 demonstration must supply both technical proof and a credible path to ultra-low-cost heavy lift.
Source: arstechnica.com
More from Marcus Bennett
Show full bio
Senior reporter covering companies, deals and strategy in artificial intelligence and machine learning at AI In Context.
791 articles
Related articles
- Google Taps Blue Origin's New Glenn to Launch AI Data Centers Into Orbit
- Google Books New Glenn to Launch Its Orbital AI Data Centers in 2029
- Google's Suncatcher sends a fridge-sized AI satellite to orbit on October 1
- Google Will Launch Its First AI Test Satellite on Oct. 1
- Google will launch AI chips into orbit on October 1st