🔍 Read the full analysis: A Haiku In Space: What Google’s Satellite Says About AI’s Possibilities on ThorstenMeyerAI.com
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
TL;DR
Google says Gemma 4 generated a haiku aboard a satellite carrying a Google TPU, marking a small-scale test of its Project Suncatcher research. The result shows that an AI chip can run in orbit, but the satellite’s limited power and unresolved costs and engineering challenges do not establish that orbital data centers are practical.
Google says its Gemma 4 model generated a haiku aboard a satellite carrying one of the company’s TPUs, marking a first operating test for its Project Suncatcher research into AI computing in orbit. The small demonstration shows that the hardware can run in space; it does not show that orbital data centers are ready to compete with facilities on Earth.
The satellite launched on October 1, 2026, on a SpaceX rideshare mission, according to Google. Built with satellite company Planet, it spent about three weeks undergoing checks before engineers sent it a prompt during a pass over Svalbard, Norway. Google said the satellite returned a three-line poem: “Blue marble below / Data streams in silent grace / Starlight feeds my core.”
The spacecraft carries four TPUs, but its available power is about one kilowatt, TechCrunch reported. That limit, along with heat constraints, means it can run only one chip at a time and only in short bursts. Google described the initial interaction as its “Hello World” moment; engineers also tested space-themed limericks and asked the payload to report its technical status.
Google calls this the first AI response from a data-center-class TPU in orbit. That wording is narrower than saying AI has never run in space: CNBC reported that startup Starcloud launched a satellite with an Nvidia H100 in November 2025 and later ran and queried Google’s Gemma model in orbit. Google’s stated distinction is its use of a TPU and its longer-term plan to connect satellites into computing clusters.
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
Testing AI Beyond Earth
The test matters as a technical milestone, not as evidence that space computing can yet solve data centers’ energy needs. Google’s Project Suncatcher explores whether satellites powered by sunlight could eventually provide computing capacity as demand for AI infrastructure grows. The company says solar panels in a suitable orbit could be up to eight times more productive than on Earth, but that is a projection about energy generation, not a result established by this small satellite.
The economic hurdle remains substantial. Google’s research estimates that, at current launch prices, launching a kilowatt of computing capacity would cost about $14,700 per year. Its estimate falls to about $810 a year if launch costs reach $200 per kilogram, a level Google says would make orbital power costs closer to those paid by US data centers. That scenario would require roughly 1,800 Starship launches, according to the source material. The test offers no evidence that those launch prices or flight rates are available now.
For companies and the public, the development is an early indicator of how far technology firms may look for energy and computing capacity. It does not change current access to AI services or show that workloads are moving off Earth at commercial scale. The immediate value is in testing equipment and gathering data about conditions that terrestrial data centers do not face.
Products Worth Considering
Jackery Explorer 2000 v2 Portable Power Station, 2042Wh
Anker SOLIX C300 Portable Power Station, 288Wh
Bose SoundLink Flex (2nd Gen) Portable Bluetooth Speaker, Black
As an affiliate, we earn on qualifying purchases.
From Research to Orbit
Google introduced Project Suncatcher in a research post and preprint on November 4, 2025. The proposal described solar-powered satellites carrying TPUs, flying close together and using laser links to share data. Google initially identified a learning mission with Planet, involving two prototype satellites, as a milestone it aimed to reach by early 2027.
The October 2026 flight brought a smaller, earlier test into orbit. Google said it confirmed contact with the spacecraft on launch day and that it was operating as expected. The company also published a peer-reviewed version of its research in the journal Joule on October 1. The haiku test followed commissioning work, including powering up the host computer and TPU trays and running early machine-learning tests.
Google’s paper sketches a future cluster of 81 satellites at an altitude of about 650 kilometers, with spacecraft flying 100 to 200 meters apart. The research describes a bench test of an optical link between a pair of transceivers reaching 1.6 terabits per second. Those proposed cluster dimensions and link results are part of the research, not proof that a working orbital cluster has been deployed.
““As the satellite passed by Svalbard, Norway, the downlinked poem confirmed success.””
— Google, in its October 30, 2026, account of the test
Cost and Durability Questions
The haiku test does not establish how reliably the TPU can operate over long periods in orbit. Google says it plans to test thermal stress and radiation exposure; results from those tests have not yet been reported. The satellite’s limited power also restricts it to short operating periods, leaving open how much useful computing it can perform.
Other challenges include removing heat in a vacuum, repairing or replacing equipment, and maintaining reliable links to ground stations and between satellites. The source material cites about 47,000 tracked objects in orbit, but does not specify the count’s date or tracking criteria, so it should be treated as a reported figure rather than a complete measure of collision risk. Google’s launch-cost estimates depend on much cheaper future launches, and it is not clear when, or whether, that cost level can be reached.
Thermal Tests and Twin Satellites
Google says the current spacecraft will undergo further tests of TPU performance under radiation and thermal stress. The next stated milestone is a pair of interconnected satellites in 2027. According to the source material, that follow-on mission is intended to generate more power and test optical links between spacecraft.
Those results will help show whether the system can progress beyond isolated, short-running demonstrations toward connected computing in orbit. Google has not established a commercial deployment schedule or shown that its projected launch economics have been achieved. The next evidence to watch is whether the planned tests work and whether the twin-satellite mission can demonstrate a reliable link and sustained computing under space conditions.
Key Questions
What did Google’s satellite do?
Google says its satellite carried a Google TPU that ran Gemma 4 and returned a haiku after engineers sent it a prompt during a pass over Svalbard. The company described the test as its “Hello World” moment for the payload.
Was this the first AI system in orbit?
No. The source material says Starcloud launched a satellite with an Nvidia H100 in November 2025 and later ran and queried Gemma there. Google’s more specific claim is that this was the first AI response from a data-center-class TPU in orbit.
Can the satellite run AI continuously?
Not according to the reported hardware limits. The satellite has about one kilowatt of power and can operate one of its four TPUs at a time in short bursts. Google has not reported continuous operation.
Why put AI computing in space?
Project Suncatcher explores whether solar-powered satellites could provide computing capacity as demand for AI grows. Google’s research argues that solar panels in a suitable orbit could be more productive than on Earth, but the cost and engineering case remains unresolved.
What is Google testing next?
Google says it will test the TPU’s response to radiation and thermal stress. Its next stated milestone is a pair of interconnected satellites in 2027 to test higher power and optical links between spacecraft.
Source: ThorstenMeyerAI.com
