# Starcloud

> Starcloud is a US deep-tech company building data centers in orbit. By placing GPU clusters in space, it taps near-continuous solar power and passive radiative cooling to run AI training and inference without the land, water, grid-permitting, and energy constraints that bottleneck data centers on Earth. In November 2025 its Starcloud-1 satellite became the first spacecraft to train a large language model in orbit using an Nvidia H100 GPU; its $170M Series A in March 2026 made it the fastest unicorn in Y Combinator history.

- **Founded:** 2024
- **Headquarters:** Redmond, Washington, United States
- **Founders:** Philip Johnston (Co-Founder & CEO), Ezra Feilden (Co-Founder & CTO), Adi Oltean (Co-Founder & Chief Engineer)
- **Team size:** ~26 employees
- **Products:** Starcloud-1, Starcloud-2, Orbital data centers
- **Notable:** First entity to train a large language model entirely in space (November 2025, Starcloud-1)., First to operate an Nvidia H100 in orbit - ~100x more powerful than any GPU previously run in space., Ran Google's Gemma model as an in-orbit chatbot, which messaged Earth from space.

## Products & services

- **Starcloud-1** — First satellite, launched November 2025. Refrigerator-sized, ~130 lbs, carrying a single Nvidia H100 GPU with solar panels, radiation shielding, and ISS-derived heat rejection. Became the first spacecraft to train an LLM in orbit (NanoGPT on the works of Shakespeare) and ran Google's Gemma model as an in-orbit chatbot.
- **Starcloud-2** — Next-generation satellite slated to launch before the end of 2026, primarily funded by the Series A. ~100x the power generation of Starcloud-1 and carrying Nvidia's Blackwell B200, one of the most powerful AI chips available.
- **Orbital data centers** — Long-term plan to scale up gigawatt-class GPU clusters in low Earth orbit, powered by near-continuous solar energy and cooled passively by radiating heat into space.

## Achievements

- First entity to train a large language model entirely in space (November 2025, Starcloud-1).
- First to operate an Nvidia H100 in orbit - ~100x more powerful than any GPU previously run in space.
- Ran Google's Gemma model as an in-orbit chatbot, which messaged Earth from space.
- Fastest company to reach unicorn status in Y Combinator history - $1.1B valuation 17 months after demo day.
- Launched its first satellite roughly 21 months after founding.

## Latest updates

- **2026-03** — Closed $170M Series A led by Benchmark and EQT Ventures at a $1.1B valuation, becoming the fastest YC unicorn ever; reportedly seeking additional funding shortly after.
- **2025-12** — Announced first-ever in-space AI model training on Starcloud-1's Nvidia H100; ran Gemma as an orbital chatbot.
- **2025-11** — Launched Starcloud-1, its first satellite, carrying a single Nvidia H100 GPU.
- **2025-03** — Rebranded from Lumen Orbit to Starcloud after a trademark challenge from Lumen Technologies and raised ~$10M more in seed funding.

## Links

- Website: https://starcloud.com
- LinkedIn: https://www.linkedin.com/company/starcloudinc/
- Twitter/X: https://x.com/Starcloud_

---

Profile page: https://yespress.io/starcloud
Published by YesPress — https://yespress.io
Last updated: 2026-06-12