🔍 Read the full analysis: Google’s Satellite Wrote A Haiku: 5 Questions About AI In Space on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Google says a prototype satellite carrying its TPU chip ran Gemma 4 and returned a haiku during a ground-station pass this week. The test marks an early step in Project Suncatcher, but the satellite’s limited power and unresolved costs, heat and radiation challenges leave large-scale orbital AI unproven.
Google says a satellite carrying one of its data-center-class TPU chips generated a haiku with its Gemma 4 AI model and sent the result to a ground station this week. The test is the first reported prompt response from Google’s Project Suncatcher, an effort to study whether solar-powered satellites could eventually perform AI computing in orbit.
The prototype launched on October 1 aboard a SpaceX rideshare mission, alongside more than 100 other spacecraft, according to the source report. Google said it spent about three weeks commissioning the system, including powering up the host computer and TPU hardware and running early machine-learning tests. The company reported that the satellite had made contact and was operating as expected.
During a pass over Svalbard, Norway, engineers sent a prompt asking Gemma 4 to write a haiku about space. Google said the satellite returned: “Blue marble below / Data streams in silent grace / Starlight feeds my core.” Engineers also tried space-themed limericks and asked the system to report its technical status. A Google spokesperson told TechCrunch that the satellite was “operating nominally.”
The spacecraft carries four TPUs but produces up to about one kilowatt of power, TechCrunch reported. That limit means it can run only one chip at a time, in short bursts. The source account says Google described the latest run as lasting about 20 minutes; it does not establish that the system can sustain AI workloads continuously.
Infographic 01 · What flew and what it did
One satellite, four chips, one haiku
Google’s Project Suncatcher prototype is the first test of whether its AI data center chips can work in orbit. Here is what is on board and how it got there.
- 4 Nov 2025Project Suncatcher announcedResearch post and preprint; plan: two prototypes with Planet by early 2027
- 1 Oct 2026Prototype launchedSpaceX Transporter-18 rideshare, built with Planet; peer-reviewed paper in Joule
- Oct 2026About three weeks of commissioningHost and TPU trays powered up, early machine learning tests
- Week of 26 Oct“Hello World”First prompt to Gemma 4; the haiku came down during a pass over Svalbard, Norway
- NextHeat and radiation tests in orbitHow the TPUs behave under thermal stress and real space radiation
- 2027Twin interconnected satellitesMore power and laser links between satellites (Google, TechCrunch)
Scale check
The satellite vs. one power deal on Earth
890 MW = 890,000 kW, about 890,000 times the prototype (my arithmetic). The prototype bar is drawn at a visible minimum; to scale it would be invisible.
Not the first AI chip in orbit
Starcloud-1, launched November 2025 (CNBC)
- Carried one Nvidia H100, a data center GPU
- Trained a small language model in orbit
- Ran Google’s open model Gemma up there in December 2025
- Google’s claim is narrower: the first response from a data center class TPU in orbit
Sources: Google, 30 Oct 2026 · Google, 1 Oct 2026 · Google Research, 4 Nov 2025 · TechCrunch, 30 Oct 2026 · TechCrunch, 1 Oct 2026 · Google and Constellation, 6 Oct 2026 · CNBC, 10 Dec 2025.
Powered by Thorsten Meyer AI
A Small Test of Orbital AI
The demonstration shows that a Google TPU can run an AI prompt in orbit and return an answer, but it does not show that a satellite data center is ready to compete with facilities on Earth. The prototype’s power supply and short operating windows put it far from the scale required for large training jobs or continuous commercial service.
Google’s larger proposal is to connect groups of satellites with laser links so they could work together like a computing cluster. That approach could matter if orbital solar power and communications can support energy-intensive AI workloads. It would also move some computing infrastructure beyond terrestrial data centers, whose electricity demand is drawing increasing attention. The test establishes a hardware milestone; it does not establish lower costs, reliable service or environmental benefits.
Popular right now
De'Longhi Magnifica Duo Fully Automatic Espresso Machine with Hot and Cold Milk Frother
Philips Baristina Espresso Machine with Milk Frother (BAR308/60)
Dyson Ball Animal 3 Upright Corded Vacuum
As an affiliate, we earn on qualifying purchases.
From Project Plan to Orbit
Google introduced Project Suncatcher in November 2025, describing a long-term concept involving solar-powered satellites equipped with TPUs and linked by optical communications. Its earlier stated goal was a learning mission with Planet, the satellite company, involving two prototypes by early 2027. The single-satellite test launched in October 2026, ahead of that stated target, but is smaller than the interconnected system proposed for later tests.
The haiku should not be described as the first AI response from space. According to the source material, Starcloud launched a satellite with an Nvidia H100 GPU in November 2025, and CNBC later reported that it ran and queried Google’s Gemma model in orbit. Google’s more specific claim is that its test produced the first response from a data-center-class TPU in orbit. Project Suncatcher’s distinctive ambition is to link multiple satellites for shared computing, rather than operate AI on a single spacecraft alone.
Google’s research, published in the journal Joule on October 1, considers tightly spaced satellite clusters and laser links. The company has reported a bench test reaching 1.6 terabits per second with a pair of optical transceivers. That figure describes a ground test, not a demonstrated satellite-to-satellite link in orbit.
““As the satellite passed by Svalbard, Norway, the downlinked poem confirmed success.””
— Google, describing the prompt test
Costs and Orbital Limits Remain
Commercial viability remains unproven. Google’s research estimates annual launch costs of about $14,700 per kilowatt at current prices, compared with about $810 per kilowatt if launch costs fall to $200 per kilogram. The lower estimate depends on a major fall in launch costs; it is not a current operating cost. The source says reaching that price case would require roughly 1,800 Starship launches. The timing, funding and feasibility of such a launch rate are not established here.
Other unresolved issues include managing heat in a vacuum, protecting chips from radiation, making repairs to hardware in orbit, and maintaining high-capacity links to the ground. The source also cites about 47,000 tracked objects in orbit, but gives no date or definition for that count. The prototype has not demonstrated a connected satellite cluster, sustained operation at data-center scale, or performance over an extended period. Those outcomes remain open questions.
Figures about terrestrial electricity use in the source are projections and estimates, not measurements of this satellite’s impact. The source reports U.S. data centers used about 4.4% of national electricity in 2023 and cites a Department of Energy forecast of 6.7% to 12% by 2028. Those figures explain interest in alternative power sources; they do not show that orbital computing will reduce electricity use or costs.
Thermal Tests Before Two Satellites
Google says it will test how the TPU hardware performs under thermal stress and radiation exposure. Those results should help show whether the chips can operate reliably in the space environment, beyond the brief prompt demonstration reported so far.
If the tests go well, Google’s next stated milestone is two interconnected satellites in 2027. TechCrunch reported that the pair is intended to generate more power and test optical links between spacecraft. That mission would address a central part of Suncatcher’s concept, but the source does not establish a launch date or confirm that the mission will proceed on schedule.
Key Questions
What did Google’s satellite do?
Google says it ran its Gemma 4 model on an orbiting TPU after engineers sent a prompt during a pass over Svalbard. The satellite returned a haiku to a ground station.
Was this the first AI response from space?
No such broad claim is supported by the source material. Starcloud had previously launched an Nvidia H100 and reportedly ran Google’s Gemma model in orbit. Google’s claim is specifically about a response from a data-center-class TPU in orbit.
Can the satellite run a full data center workload?
The reported prototype produces up to about one kilowatt and can operate one of its four TPUs at a time in short bursts. It has not demonstrated sustained, data-center-scale computing.
Why put AI chips in orbit?
Google is investigating whether satellites with solar panels could provide power for computing and use laser links to connect into clusters. Whether that can be done reliably and at a competitive cost is still unknown.
What is the next milestone?
Google says it will test the TPU’s response to heat and radiation. If those tests go well, the company’s next stated goal is a pair of interconnected satellites in 2027.
Source: ThorstenMeyerAI.com
