🤖🦾 The Need For Sovereign Compute

A Newsletter for Entrepreneurs, Investors, and Computing Geeks

Author

Benjamin Wolba
March 17, 2025

It’s been a packed week in computing. Analog quantum computers have made headlines, but classical supercomputers might not be conceding defeat just yet. And the EU is doubling down on sovereign compute with six new AI factories.

Across the Atlantic, all eyes are on GTC in San José this week, where we expect major announcements shaping the AI and computing landscape.

And speaking of events, our London conference is completely booked out—but don’t worry, we’re already gearing up for the next one in Berlin on June 2 (📍RSVP here), with the first speakers to be announced soon.

Plus, if you're in Berlin on April 23, join us for the next Future of Neuromorphic Computing Meetup (sign up here).

Spotlights

“Researchers have sought materials that can switched between dark and bright states when exposed to light that could make switches for future photonic computers. Now scientists in California say they have made nanoparticles with this property for the first time.”

“With the exception of custom cloud silicon, like Google's TPUs or Amazon's Trainium ASICs, the vast majority of AI training clusters being built today are powered by Nvidia GPUs. But while Nvidia may have won the AI training battle, the inference fight is far from decided.”

“Despite bans put in place by the U.S. and Europe on the supply of electronic components to Russia, dozens of Russian microelectronics suppliers continue to obtain and resell imported chips to Russian arms manufacturers successfully.”

“In just minutes, a special quantum processor, called a quantum annealing processor, solved a complex real-world problem that a classical supercomputer would take millions of years to complete, researchers claim March 12 in Science. ... However, another group of researchers claims to have already found a way for a classical supercomputer to solve a subset of the same problem in just over two hours.”

Headlines

⚡️ Chinese researchers develop ultra-high-capacity chip, facilitating on-chip optical data transmission (Global Times)

⚡️ Energy-Efficient Scalable Silicon Photonic Platform For AI Accelerator HW (SemiconductorEngineering)

⚛️ Pasqal Quantum Available on Azure (insideHPC)

Funding News

⚛️ $9M Seed – QuamCore: quantum computing scaling technology built for superconducting quantum processor architectures (SiliconAngle)

🦾 €13.7M Series A+ – ATLANT 3D: building an advanced atomic-scale manufacturing platform for the semiconductor industry

⚡️ $250M Series C – Celestial AI: optical computing and interconnect technologies (Photonics)

Bonus:

Source: EU Commission

Deep Dive: EU Striving for Sovereign Compute

Europe is increasingly recognizing the importance of establishing sovereign artificial intelligence (AI) computing capabilities to maintain technological independence and drive innovation.

For example, Denmark has launched its first AI supercomputer, Gefion. Operated by the Danish Center for AI Innovation (DCAI), Gefion aims to tackle challenges in quantum computing, clean energy, and biotechnology, serving as a foundation for Denmark's sovereign AI initiatives.

In Finland, the LUMI supercomputer—one of the most powerful supercomputers globally—is a testament to Europe's commitment to high-performance computing.

And just last week, the European High Performance Computing Joint Undertaking (EuroHPC JU) has selected six additional AI Factories in Austria, Bulgaria, France, Germany, Poland, and Slovenia, supported by a combined investment of approximately €485 million. These facilities will provide startups and small-to-medium enterprises with access to advanced AI resources, fostering growth and innovation across Europe.

These developments underscore Europe's strategic efforts to build sovereign AI computing infrastructures, ensuring control over critical technologies and data.

💡 Love these insights? Forward this newsletter to a friend or two who can subscribe here