08 · SPACE

Electronics in Space

Inference chips optimized for mass, thermal efficiency, and radiation tolerance for space compute.

Championed by Philip Johnston at YC

THE PROBLEM

What needs to be solved

Space electronics are stuck in the 2000s. The radiation-hardened chips used in most satellites today are 3-5 generations behind consumer hardware — slower, more expensive, and incapable of running modern AI models. A satellite that could process imagery on-orbit with current AI chips would eliminate the need to downlink terabytes of raw data, but no radiation-tolerant chip exists that can run inference workloads efficiently.

WHY NOW

What changed in 2025–2026

The number of satellites launched annually grew from 200 in 2018 to over 2,500 in 2025. SpaceX's Starship dramatically reduces launch costs, making larger constellations economically viable. Meanwhile, defense and commercial satellite operators desperately need on-orbit AI processing for real-time Earth observation, communications optimization, and autonomous operations. The demand exists; the hardware doesn't.

MARKET CONTEXT

The size of the opportunity

The space electronics market is projected to reach $5 billion by 2028. Traditional rad-hard chip makers (BAE Systems, Microchip) move slowly and charge 100-1000x consumer chip prices. New approaches — radiation-tolerant chip architectures, FPGA-based solutions, and novel packaging — can deliver 10x performance at 1/10th the cost. The DoD and commercial constellation operators (Planet, Spire, Amazon Kuiper) are actively seeking suppliers.

FOUNDER FIT

Who should build this

This requires deep semiconductor expertise — founders who understand chip architecture, radiation effects, and space-grade qualification processes. Ideal backgrounds include ASIC/FPGA design, radiation testing, or satellite systems engineering. Teams from companies like SpaceX, Planet Labs, or traditional chip companies with space division experience are well-positioned. This is a hardware play, so expect longer development cycles and higher capital requirements.

WHAT YC SAYS

The YC partner perspective

Philip Johnston identifies this as a critical infrastructure gap. As space becomes more commercialized and AI becomes essential for satellite operations, the demand for space-grade inference chips will explode. The founders who solve this will own a chokepoint in the space economy — every satellite constellation will need their chips.

GO DEEPER

Get the complete Electronics in Space playbook

The full playbook includes an 8-week MVP plan, pricing model with unit economics, competitor analysis, customer acquisition strategy, risk mitigations, and a day-by-day 90-day action plan to get to first revenue.

Get the Full Playbook — $49 →