⚛️🦾 What Makes Qubits Topological?

A Newsletter for Entrepreneurs, Investors, and Computing Geeks

Another exciting week in computing. xAI launched Grok 3, an AI model that competes with highlighting the power of scale over heuristics, and Microsoft's Majorana 1 quantum processor made headlines, claiming a breakthrough in topological quantum computing.

We were busy preparing for the Future of Computing Conference 2025 in London, which will take place on March 26 this year, and we’re. We’re looking forward to seeing some of you there!

Spotlights

Grok has just released Grok 3, an AI model that rivals or surpasses those from established labs like OpenAI and Google DeepMind. This achievement underscores the "Bitter Lesson," which posits that scaling compute power is more effective than relying on heuristic solutions. While optimization has its place, access to substantial computational resources, as demonstrated by xAI's Grok 3, is the more decisive factor in AI progress.

“In the late 1950s, Sweden briefly had the world’s fastest computer. And then in the 1960s, they produced computers good enough to beat even the vaunted IBM. An amazing start. Sweden might have been the computing leader of Europe. But unfortunately, that did not last. In this video Asianometry looks at the birth and glory of Swedish-made computers with the story of Datasaab.”

“Below is my attempt to build such a capable AI computer without spending too much. I ended up with a workstation with 48GB of VRAM that cost me around 1700 euros.”

“Cluster deployments are an order of magnitude larger in scale, with Gigawatt-scale datacenters coming online at full capacity much faster than most believe. As such, there are considerable design changes that data center developers planning future sites must consider.”

“In 2014, Loff and four other researchers discovered that adding full storage space can, in principle, make computers more powerful. Their theoretical framework, called catalytic computing, has become an object of study in its own right.”

Headlines

🤖 Open source LLMs hit Europe’s digital sovereignty roadmap (TechCrunch)

🤖 Thinking Machines Lab is ex-OpenAI CTO Mira Murati’s new startup (TechCrunch)

🤖🌐 On-Device Machine Learning in Spatial Computing (Towards Data Science)

🤖🦾 ZeroPoint Technologies Unveils Groundbreaking Compression Solution to Increase Foundational Model Addressable Memory by 50% (Zero Point)

🤖⚛️ Quantum generative AI: Quantinuum CEO Raj Hazra sat down with UKTN to explain why generative AI is about to get a lot more powerful (UKTN)

⚛️🛡️ Netherlands Ministry of Defence Launches Quantum Challenges with Six Industry Partners (Quantum Insider)

⚡️ Hyperlume wants to make chip-to-chip communication faster and more efficient (TechCrunch)

⚡️ Advanced Packaging Evolution: Chiplet And Silicon Photonics-CPO (SemiconductorEngineering)

⚡️ STMicro, AWS co-develop AI-focused photonic chip for data centers (DigiTimesAsia)

⚡️🔊 Scalable way to generate and control sound in silicon photonic chips (Phys Org)

🧠 Leveraging stochasticity in memristive synapses for efficient and reliable neuromorphic systems (Nature)

💥 How Intel ruined an Israeli startup it bought for $2B—and lost the AI race (CTECH)

🚨 Trump says semiconductor tariffs will start at 25%, keep rising over the year (DCD)

🦾 Arm is launching its own chip this year with Meta as a customer (TechCrunch)

🦾 NXP Agrees to Acquire Edge AI Pioneer Kinara to Redefine the Intelligent Edge (GlobeNewswire)

🦾 Germany supports the construction of a new chip factory in Dresden with 920 million euros (deutschland)

🦾 Ifo study: Germany is one of the top nations in the semiconductor industry (Heise)

🦾 Meta’s Hyperscale Infrastructure: A look at Meta's planetary-scale computing infrastructure (CACM)

🦾 OpenAI partners with TSMC to manufacture custom AI chip, with plans for future iterations (Ars Technica)

Funding News

⚡️ $12.5M Seed RoundHyperlume: high bandwidth, low latency, low power interconnects for data centers and high-performance computing systems (Photonics)

⚛️ €17.5M EIC fundingSemiQon: cryogenic CMOS technology (Arctic Startup)

🦾 $21.5M SeedAheadComputing: RISC-V chips (Reuters)

🤖🦾 $23.5M SeedPositron: energy-efficient AI chips (Verdict)

⚛️ $28M Series AQuSecure: post-quantum cryptography (PQC) solutions (Quantum Computing Report)

⚡️ $30M Series ASalience Labs: photonic solutions for AI data center infrastructure connectivity (Finsmes)

🦾 $100M Series BEnCharge AI: analog in-memory chips (DCD)

⚛️ $230M Series BQuEra Computing: neutral-atom quantum computing (Finsmes)

Bonus Material for Researchers

Welcome to the Deep Tech Fellowship, a paid, three-month part-time program in Munich that connects PhD students with the deep tech startup and VC ecosystem.

6-month, part-time, remote program for PhD candidates and PostDocs to explore emerging scientific domains with breakthrough potential

Deep Dive: Topological Quantum Computing

It’s been all over the news these days: Microsoft has announced Majorana 1, the world's first quantum processor powered by topological qubits.

Unlike conventional qubits, topological qubits are designed for greater stability and error resistance. They rely on topological states of matter, where quantum information is distributed across the system rather than confined to individual particles—making them less susceptible to noise.

If you’re curious about what topological qubits are and whether Microsoft has truly cracked quantum computing, Scott Aaronson’s blog is a must-read.

Also worth noting: Nokia Bell Labs just revealed that they, too, have been working on topological quantum computing for years.

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