German semiconductor startup Ubitium has raised $14 million in seed funding after producing working first silicon for a processor it says could replace several specialized chips inside robots, drones and other autonomous machines.
The Düsseldorf-based company said the round was led by Ibex Investors, with participation from OpenOcean, KBC Focus Fund, Runa Capital, Inflection and Florian Seibel, co-founder and co-CEO of European drone company Quantum Systems. The financing brings Ubitium’s total seed funding since its 2024 founding to $18 million.
Independent European startup publication Startup.eu reported that Ubitium has now reached the important milestone of working first silicon while raising the new capital to continue development of its universal processor.
One chip for perception, inference and control
Ubitium’s core thesis is that autonomous machines currently rely on too many different processors.
The company says physical AI systems generally need computing for signal processing, AI inference and real-time control. Instead of dividing those workloads across different chips, Ubitium is developing one universal processor intended to handle all three categories.
Startup.eu describes the architecture as an attempt to replace the mix of specialized chips typically found in autonomous machines such as industrial equipment, drones and robots.
That could matter because every additional processor adds integration work, software dependencies, power requirements and engineering complexity.
Ubitium is basing the project on RISC-V, the open instruction-set architecture increasingly attracting interest as an alternative to proprietary processor ecosystems.
First silicon changes the nature of the startup’s bet
Chip startups face a much harder validation process than software companies.
A design can appear promising in simulation but still encounter manufacturing, power or performance problems when physical silicon arrives.
Ubitium said it reached its first-silicon milestone in July 2026 and is now moving from architectural claims toward hardware that developers can eventually test against real workloads.
For investors, that reduces one layer of technical uncertainty while creating the next challenge: proving that a general-purpose architecture can compete with highly optimized processors built for individual tasks.
Europe is investing beyond AI software
The funding also points to a broader European technology shift.
Much of the current AI startup market is concentrated around software, models and applications. Physical AI requires a deeper stack involving processors, sensors, robotics software and real-time control systems.
Europe already has significant industrial, robotics and automotive expertise. Ubitium is betting that those markets need a different computing architecture as machines become more autonomous.
The idea is ambitious because specialized chips exist for a reason: narrowly designed hardware can deliver exceptional performance and efficiency.
Ubitium’s opportunity is proving that reducing system complexity can outweigh some of those advantages.
For founders and engineers, that makes this $14 million round more interesting than the funding number itself. The startup is testing whether the next physical-AI platform should contain more accelerators—or fewer chips capable of doing much more.
