The Fermi Explorer interstellar mission asks a deceptively simple question: what if the point was never to get there quickly? Backed by the founders of orbital data-centre startup Starcloud, the newly announced non-profit wants to send a one-kilogram spacecraft on a journey to Alpha Centauri, accepting an 80,000-year travel time rather than waiting for propulsion technology that may never materialise.
The mission emerged publicly this week, soliciting partners, donations, and scientific payloads. The core team: Starcloud co-founders Philip Johnston, Adi Oltean, and Ezra Feilden, joined by programme manager Garret Jameson. Their plan is to launch a vehicle carrying a 10-cm-square payload on a rocket rideshare in 2029, at an estimated cost of about $15 million.
The Fermi Explorer Interstellar Mission’s Low-Tech Logic
Most interstellar proposals collapse under the weight of what they cannot yet build. The last serious Silicon Valley attempt, Breakthrough Starshot, was founded in 2016 by Yuri Milner and backed by Mark Zuckerberg with a vision of laser-propelled craft reaching Alpha Centauri in twenty to thirty years. It folded in 2025.
‘I also got excited and then disappointed!’ Johnston said of Starshot. ‘That’s the reason we are doing the mission. We are hell-bent on actually launching something in three years … the big difference with most other [interstellar] missions and the Fermi Explorer is that the others all try to do it in a human lifetime, which means billion-dollar propulsion R&D programs. We have very deliberately picked 80,000 years as the minimum viable mission.’
The distances involved make any timetable humbling. Voyager, launched in 1977, sits roughly 26 billion km from Earth. Alpha Centauri, the nearest star system, is 41 trillion km away. The plan is to use electric propulsion to slingshot the spacecraft through a series of increasingly fast orbits, including close passes of approximately 0.42 astronomical units from the Sun to gain speed near perihelion. Cosmic Log reports the journey duration at approximately 77,000 years and the target as at least 99% of the distance to Alpha Centauri; the team has publicly cited 80,000 years. Both figures describe the same essential ambition.
The spacecraft is expected to weigh between 100 and 200 kg and reach a cruise velocity of approximately 23.6 km/s, remaining active for about 12 years before entering a coast phase. Johnston is sanguine about what happens after that. ‘If a dinosaur bone can last 200 million years then a metal box can easily survive the radiation environment we expect for 50,000 years. We don’t plan on powering spacecraft up at the end of the 80,000 years. We hope that a future human civilization will intercept it.’
The mission draws its name from the Fermi paradox: the unresolved tension between the statistical likelihood of extraterrestrial intelligence and our failure to detect any sign of it. Beyond the philosophical framing, the team wants to carry scientific and artistic payloads, much as NASA’s Voyager carried sensors and its Golden Record.
The Company Behind the Mission
Starcloud itself is moving at a rather different pace. The Surrey-born Johnston co-founded the startup in 2024 alongside Oltean, a former SpaceX engineer, and Feilden, an Imperial College London graduate. Yahoo Finance has reported that the company has filed with the FCC for a constellation of 88,000 spacecraft.
The commercial trajectory has been rapid. Starcloud raised a $170 million Series A at a $1.1 billion valuation, led by Benchmark and EQT Ventures, becoming Y Combinator’s fastest-ever unicorn just 17 months after completing the seed accelerator, according to a Business Wire announcement. A subsequent $250 million Series A extension, led by Manhattan West with participation from NVIDIA, Cisco Investments, and others, pushed the post-money valuation to $2.3 billion and total capital raised to $450 million, as confirmed in a separate Business Wire filing.
The company launched its first satellite, Starcloud-1, in November 2025, carrying an NVIDIA H100, described as roughly 100 times more powerful than any GPU previously in orbit. The satellite became the first to train an AI model in space. CNBC reported that the model trained was Google’s Gemma, which Johnston described as proof that orbital data centres can run a variety of AI workloads, particularly those demanding large compute clusters. Starcloud and NVIDIA are also collaborating on the NVIDIA Space-1 Vera Rubin Module, announced in March 2026.
The Fermi Explorer sits entirely outside that commercial operation, structured as a separate non-profit. Whether the founders can secure $15 million in donations and scientific interest will be the first test of whether a slow-burn, low-tech pitch can travel as far as the bolder, pricier visions that preceded it.
