Skip to content
World clockEU--:--UK--:--USA--:--CN--:--PLDEFRIT中文EN

portal about AI and technologyevents · analysis · interviews · technical background

Search
LIVE
ExplainerWearables and AI dominate a busy week in digital health research02.10AnalysisUS Order Pushes Grid Supply Chain Scrutiny Into Software Layer02.10AnalysisThe humanoid hand fight moves to a factory floor that is already full02.10AnalysisSuperfluid qubit design, a DOE roadmap and a universal gate set02.10NewsSteam Deck 2 chip rumors meet a hard datacenter bottleneck02.10NewsSony starts rolling out QSSR AI upscaling on the base PS502.10AnalysisSmall models move to the edge as Google locks down its biggest one02.10ExplainerShopify Canvas Puts an AI Chat at the Centre of Store Building02.10AnalysisShopify Canvas Lets Merchants Build Stores by Chat, as AI Agents Breach a Security Nonprofit02.10AnalysisScience Journal Turns to Women's Health as Funding Cuts Threaten the Field02.10
›

Google sends its first AI chips into orbit on a SpaceX rocket

On 1 October, a Falcon 9 is set to place four AI processors into orbit for the Suncatcher project. The prototype will compute only in sessions of about fifteen minutes, because of cooling constraints.

Rachel NwosuRachel NwosuAI, models and technology
TechnologyNewsRachel NwosuPublished: 25 September 20265 min readSources 3
Google sends its first AI chips into orbit on a SpaceX rocket

After the broad principles, the demonstration. On Thursday 1 October, Google will send four of its artificial intelligence chips into orbit aboard a SpaceX Falcon 9, Numerama reports. It is the first real-world test of the Suncatcher project, presented in November 2025, which aims to install computing centres in space.

Fifteen minutes of computing per session

The company announced the launch in a status update on 24 September. A blog post signed by Travis Beals, a director at the group, then confirmed it. The prototype on board will compute only in sessions of about fifteen minutes. In the vacuum of space, heat dissipation relies on radiation rather than convection, and that caps how long the chips can run.

The strategic goal is more ambitious than the technical demonstration: to build space data centres able to take advantage of near-continuous sunlight, with no night and no clouds, to power AI computing loads. Abundant, steady energy is precisely what ground-based centres lack most.

Why the energy question is shifting

The reasoning sits in a context of strain on electricity and land. In Italy, industry analysts note that next-generation racks, at the megawatt scale, are pushing towards direct-current architectures. Those designs cut losses and carry energy efficiency up to 98%. Solid-state transformers and smart flow management will be decisive in supporting the growth of computing capacity. The migration will happen in stages, from retrofit to hybrid architecture and then native, with challenges of standards, skills and complexity. The first prototypes and solid-state transformers are expected in 2026 and 2027.

In space, the constraint is no longer access to energy but the removal of heat: computing has to happen in short sessions.

Nothing guarantees that orbital computing will one day become competitive against ground-based centres. The test on 1 October will mainly serve to measure how the chips hold up to radiation and extreme thermal swings, before any economic conclusion.

Comments 0

Sources

3
  1. 01Numerama : Google envoie ses premières puces IA en orbite à bord d'une fusée SpaceXFR
  2. 02Ars Technica : Google's first Suncatcher orbital data center test launches October 1EN
  3. 03Punto Informatico : Google lancia i chip AI nello spazio, cos'è Project SuncatcherIT

All figures and quotations in this text come from the sources listed below.

Content prepared by the editorial team with AI assistance.

Rachel Nwosu

Rachel Nwosu

AI, models and technology

Rachel Nwosu covers AI, models and technology for FLASH24, working from public model documentation, benchmark releases and repository histories rather than press summaries, and she skips announcements that arrive without reproducible numbers. She checks training-data claims against dataset cards and reruns reported metrics where code is available. She spends much of her week interviewing researchers and engineers, tracking model launch calendars, and comparing vendor benchmarks with independent evaluations. Outside the desk she runs 3D printers, restores old computers, and tests how models learn from internet junk. She does not publish benchmark figures she cannot trace to a source.

Newsroom →

Comments

0
  1. No comments yet — be the first.

Write a comment

Comments are public. We do not publish abuse, spam or advertising.

Previous articleChina puts almost 490 billion euros into networks and computing, targeting 9,800 EflopsNext articleCybersecurity: an OpenAI agent breached Australia's public health insurance system