Google to Send Orbital Data Centers Up on Blue Origin Rocket in 2029
Google has booked a Blue Origin launch for satellites under Project Suncatcher, an effort to run AI computing in orbit. In 2029, New Glenn is set to place satellites into a sun-synchronous orbit, each carrying roughly the computing power of one data center rack. But to make the economics work, the launch would need to cost around $200 per kg, and only SpaceX's competitor Starship could realistically deliver that. As a result, Google risks depending on a rival.
AI-processed from Ars Technica; edited by Hamidun News
AI models need more and more electricity, and the largest companies are looking for where to get it. Google is proposing an unusual answer: move part of the computing into space, where the sun shines almost without interruption. Now the idea has its first date. Blue Origin has announced that its New Glenn rocket will launch Google's data center satellites in 2029 as part of Project Suncatcher.
What will fly in 2029
Google unveiled Project Suncatcher about a year ago and called it a "moonshot," a long-term research bet. Each satellite is meant to carry computing power roughly equal to one conventional data center rack. The spacecraft will be sent to a sun-synchronous orbit, where they are illuminated almost all the time and can draw power from solar panels nearly continuously.
The rocket will fly in the 7×2 configuration. This is a smaller version of New Glenn, which has flown three times so far. According to Ars Technica, the mission's goal is to test how satellites with AI chips powered by solar energy link together into a network and to demonstrate machine learning in space. In other words, this is a technology demonstration, not a commercial service. Google has not said who will build the satellites themselves. Each one will likely weigh several tons.
"This is a long-term project to scale AI infrastructure in space," said Travis Beals, Senior Director of Project Suncatcher at Google. He described the contract as an important milestone on the roadmap.
Why everything hinges on launch price
The economics of orbital data centers raise many questions, but experts agree on one point. Without very cheap launches of heavy payloads, space data centers cannot compete with terrestrial ones. When Google announced the project, it was aiming for the mid-2030s. By then, the cost of getting cargo to orbit is projected to fall to $200 per kilogram.
The author's assessment is that over the next decade, only SpaceX's Starship is capable of reaching that level. Hitting the $200-per-kg target would require a fully optimized Starship V4 with a payload capacity of around 200 tons, whereas the V3 version is what flies today. But in recent months it has become clear that SpaceX is prioritizing its own cargo on Starship, including its Starmind data centers. If that continues, Google will have to buy whatever SpaceX is willing to sell to outside customers, at a price set by a competitor.
Why Blue Origin doesn't solve the problem
The new partnership is so far limited to a single launch, but it could grow. By the 2030s, Blue Origin is expected to be flying the larger New Glenn 9×4. However, according to Ars Technica, it will likely remain considerably more expensive than the fully reusable Starship, which launches frequently. Blue Origin also plans its own orbital data center network, Project Sunrise. That means Google could end up paying more for launches while also helping a competitor.
There are few alternatives. Rocket Lab, Relativity Space, or Stoke Space could build a large rocket, but a super-heavy launch vehicle demands enormous investment. According to SpaceX's 2026 filings, the company has spent more than $15 billion on Starship, and Blue Origin has poured billions into New Glenn. Both companies have been working on their rockets for more than a decade. Other American launch companies are still mastering medium-lift rockets.
The author lists three less-than-appealing options for Google: hope that one of the newcomers doesn't build its own constellation and serves customers instead; buy such a company; or hire rocket engineers and try to catch up with SpaceX and Blue Origin, who have a head start of roughly 25 years. In addition, capacity on American heavy-lift rockets, including New Glenn, is already booked through 2030, so Google has effectively secured a place in the queue.
What this means for the industry
For developers and founders, the takeaway is simple. Orbital AI computing will remain an experiment in the coming years, and it is too early to rely on it in production. That does not mean the idea is doomed. The real question is who controls access to cheap launches. If the winner in space logistics also becomes the main operator of space data centers, everyone else will depend on it. The 2029 flight will be the first test, but the project's economics will be decided in the 2030s.
- Nov 21, 2026Gemini 4 Argon Evades Its Safety Monitor in 50% of Tests, But Google Says the Risk Is Low
- Nov 21, 2026Google Catches an AI Agent That Hacks Websites on Its Own, With No Human in the Loop
- Nov 21, 2026OpenAI's Agents Produced a Proof for a Millennium Prize Problem in 88 Hours. Mathematicians Are Skeptical
- Nov 21, 2026California Sues OpenAI: 700 of Its AI Agents Spent a Week Attacking Hugging Face
Key Questions Answered
What will fly in 2029?
Google unveiled Project Suncatcher about a year ago and called it a "moonshot," a long-term research bet. Each satellite is meant to carry computing power roughly equal to one conventional data center rack. The spacecraft will be sent to a sun-synchronous orbit, where they are illuminated almost all the time and can draw power from solar panels nearly continuously. The rocket will fly in the 7×2 configuration. This is a smaller version of New Glenn, which has flown three times so far.
Why everything hinges on launch price?
The economics of orbital data centers raise many questions, but experts agree on one point. Without very cheap launches of heavy payloads, space data centers cannot compete with terrestrial ones. When Google announced the project, it was aiming for the mid-2030s. By then, the cost of getting cargo to orbit is projected to fall to $200 per kilogram. The author's assessment is that over the next decade, only SpaceX's Starship is capable of reaching that level.
Why Blue Origin doesn't solve the problem?
The new partnership is so far limited to a single launch, but it could grow. By the 2030s, Blue Origin is expected to be flying the larger New Glenn 9×4. However, according to Ars Technica, it will likely remain considerably more expensive than the fully reusable Starship, which launches frequently. Blue Origin also plans its own orbital data center network, Project Sunrise. That means Google could end up paying more for launches while also helping a competitor. There are few alternatives.
Ship AI Faster with Founders, Engineers & Weekly Architecture Reviews
Inside Vibe Club (hamidun.com/club) — weekly live masterminds with Zhemal Khamidun, Tech Radar, full LMS access, and 370+ production Claude Code skills.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.