NVIDIA Quantum AI, Meta Broadcom Gigawatt Partnership & Amazon Satellite Acquisition

Samantha: Welcome to The State of Tech, The European Edition, Wednesday April fifteen, 2026. I'm Samantha Lawrence.

Bob: And I'm Bob Russell. Today: NVIDIA goes open source to crack the quantum computing puzzle, Broadcom and Meta team up for a multi-gigawatt AI chip marathon, Amazon's AI agents are designing antibodies at breathtaking speed, Lucid and Uber are scaling robotaxis to thirty-five thousand vehicles, Amazon makes a ten-point-eight billion dollar satellite play for Globalstar, and Intel teases its next-gen desktop chips amid a turbulent hardware market. Let's start with quantum.

NVIDIA bets that AI is the missing control layer quantum computing has been waiting for.

Samantha: So NVIDIA just dropped something pretty significant. They're calling it Ising — and it's the world's first family of open-source AI models built specifically to accelerate quantum computing. The name's a nice nod to the Ising model in physics, by the way.

Bob: And the core idea here is that quantum computers are still incredibly fragile. Qubits lose coherence, error rates are high, calibration is a nightmare. What NVIDIA is saying is: let AI handle that mess. These Ising models focus on two critical problems — quantum processor calibration and quantum error correction. And the numbers are genuinely impressive. We're talking up to two and a half times faster performance and three times higher accuracy for error correction decoding.

Samantha: That error correction piece is huge. Because without reliable error correction, you can't scale quantum. You just can't. So positioning AI as essentially the control plane for quantum machines — that's a bold architectural bet.

Bob: And the fact that it's open source changes the game. This isn't locked behind a proprietary wall. Anyone can build on it, adapt it, improve it. That's a deliberate ecosystem play by NVIDIA.

Samantha: Which brings us to Europe. IQM Quantum Computers in Finland and the UK's National Physical Laboratory are already among the early adopters. European quantum labs have been doing excellent fundamental research but often struggle with the engineering side of scaling up. This gives them a massive shortcut.

Bob: It could genuinely reduce Europe's dependence on proprietary quantum stacks controlled by a handful of American companies. If European researchers can freely build on these models, you could see faster progress in drug discovery, materials science, logistics optimization — all areas where Europe has strong academic foundations.

Samantha: There's a strategic angle too. The EU has been pouring money into quantum through the Quantum Flagship programme. Having open-source tools like Ising available means those public investments could go further and deliver results faster.

Bob: The only caution I'd add is that open source doesn't mean open playing field. NVIDIA still controls the GPU hardware these models run on. So there's a layer of dependency that doesn't disappear. But as a starting point for democratizing quantum development? This is a significant move.

Samantha: Agreed. And if AI really does become the brain that makes quantum hardware usable, NVIDIA just positioned itself right at that intersection. Smart.

---

Meta wants personal superintelligence for billions of users — and Broadcom is building the silicon to make it happen.

Bob: Next up, Broadcom and Meta have locked in a multi-year, multi-generation partnership to scale Meta's AI infrastructure. And the numbers here are staggering.

Samantha: We're talking about an initial commitment exceeding one gigawatt of compute power, with plans for a sustained multi-gigawatt rollout through 2029. To put that in perspective, one gigawatt is roughly the output of a nuclear power plant. And they want multiple gigawatts dedicated to AI.

Bob: Broadcom's role is supporting Meta's custom MTIA chips — that's Meta Training and Inference Accelerator. This isn't Meta buying off-the-shelf GPUs. This is deep co-design through Broadcom's XPU platform, where the silicon is purpose-built for Meta's workloads.

Samantha: And what's the end goal? Meta is explicitly talking about delivering what they call personal superintelligence to billions of users across WhatsApp, Instagram, and their other platforms. That phrase — personal superintelligence — should make everyone sit up.

Bob: It's ambitious language. What it means in practice is real-time generative AI features that are deeply personalized. Think AI assistants that genuinely know your preferences, your communication style, your needs. Running at scale across billions of accounts.

Samantha: For Europe, this is a double-edged sword. On one hand, European users get access to incredibly powerful AI features. European data centers will see increased investment and job creation. On the other hand, personal superintelligence running on data from billions of Europeans raises every privacy alarm the EU has been sounding for years.

Bob: The GDPR implications alone are enormous. How do you deliver deeply personalized AI without processing vast amounts of personal data? The EU AI Act adds another layer. European regulators will be watching this very closely, and I wouldn't be surprised if we see friction between Meta's ambitions and European data protection frameworks.

Samantha: It's the classic tension. Innovation at incredible speed versus regulatory caution designed to protect citizens. Neither side is wrong, but the collision is coming.

---

Samantha: Quick interruption. If you listen to The State of Tech regularly, hit that like button and subscribe — that way you'll never miss an episode. Okay, moving on.

Amazon's AI agents just did in weeks what used to take a year in drug discovery.

Bob: Amazon Biopharma Research has unveiled a platform that uses AI agents and multiple biological AI models to design antibodies. And the speed here is remarkable.

Samantha: They generated nearly three hundred thousand antibody molecule designs, selected the top hundred thousand candidates, and sent them for wet lab testing. The whole process — which traditionally takes up to a year — was done in weeks. Weeks, Bob.

Bob: The target is rare pediatric cancer, which makes this even more compelling. These are diseases where traditional drug discovery economics often fail because the patient populations are small and the financial returns are limited. If AI can dramatically cut the cost and time of development, suddenly those economics shift.

Samantha: What's technically interesting is that this isn't a single AI model doing everything. It's a network of AI agents working together — designing molecules, evaluating them, iterating. It's an AI-driven scientific workflow that integrates directly into the wet lab pipeline.

Bob: For European pharma, this should be both exciting and alarming. Exciting because European companies and research institutions could adopt similar approaches. The Karolinska Institute, the Pasteur Institute, major European biotech clusters — they could all benefit from this kind of acceleration.

Samantha: But alarming because if Amazon is setting the pace, European biopharma risks falling behind. The continent has world-class biological research, but translating that into rapid drug development has always been a bottleneck. American tech giants bringing AI-scale speed to drug discovery could widen that gap.

Bob: There's also a regulatory dimension. The European Medicines Agency will need frameworks for evaluating drugs that were designed by AI agents rather than traditional medicinal chemistry teams. The science is moving faster than the regulatory apparatus.

Samantha: The bottom line is that AI is no longer just assisting drug discovery. It's driving it. And the implications for patients — especially children with rare cancers — could be transformative.

---

Uber's robotaxi fleet is about to get a whole lot bigger — and a whole lot more luxurious.

Samantha: Lucid and Uber are deepening their robotaxi partnership in a big way. Uber is now committed to purchasing at least thirty-five thousand Lucid vehicles for its global autonomous ride-hailing service.

Bob: And the financial backing is serious. Uber has now invested a total of five hundred million dollars in Lucid. The Public Investment Fund — that's Saudi Arabia's sovereign wealth fund — is also putting in additional long-term investment. The commercial launch starts later this year in the San Francisco Bay Area using the Lucid Gravity SUV.

Samantha: Thirty-five thousand vehicles is a massive fleet. That's not a pilot programme anymore. That's an operational commitment at real scale. And using Lucid vehicles — which are premium, long-range electric cars — signals that this robotaxi service is aiming for a higher-end experience.

Bob: For Europe, the big question is when, not if. Uber already operates across major European cities. Once the US operations prove the model, expansion to London, Paris, Berlin, Amsterdam — that's the logical next step.

Samantha: But European cities will present unique challenges. Narrow streets, complex traffic patterns, aggressive cyclists — sorry, I'm thinking of Amsterdam specifically. And regulators across Europe have been cautious about autonomous vehicles on public roads. Each country, sometimes each city, has different rules.

Bob: That regulatory patchwork is a real obstacle. In the US, you can launch in one state and learn. In Europe, you might need separate approvals for every market. It could slow down deployment significantly. But the pressure from consumers who see it working elsewhere will push regulators to act.

Samantha: And there's the labour angle. European taxi unions and ride-hailing drivers will have strong opinions about thirty-five thousand robots taking their jobs.

---

Amazon is buying its way into the satellite internet race — to the tune of nearly eleven billion dollars.

Bob: Also today, Amazon has agreed to acquire satellite operator Globalstar for ten point eight billion dollars in cash and stock. That's a huge deal.

Samantha: This is about connecting consumer smartphones directly to satellite internet. So think of areas with no cell coverage — rural communities, remote regions, disaster zones — and your phone just works anyway because it's talking to a satellite overhead.

Bob: Amazon already has Project Kuiper, its own satellite constellation programme. Adding Globalstar gives them existing infrastructure, spectrum licences, and operational experience. It's a massive acceleration of their satellite ambitions.

Samantha: For Europe, the connectivity implications are significant. There are still plenty of areas across the continent — rural Spain, parts of Scandinavia, the Scottish Highlands, eastern European countryside — where broadband and even mobile coverage remain poor. Satellite-to-smartphone could genuinely bridge those digital divides.

Bob: But European telecom operators are going to be nervous. Companies like Deutsche Telekom, Orange, Vodafone — they've invested billions in terrestrial infrastructure. Now Amazon is potentially offering an alternative that bypasses all of that. And with Amazon's scale and resources, they can subsidize services in ways traditional telecoms simply can't match.

Samantha: There are regulatory questions too. Spectrum management in Europe is handled nationally and through EU coordination. Adding a major new satellite player changes that landscape. And data sovereignty — European data routing through American-owned satellites — will trigger familiar debates.

Bob: It's worth noting this also puts Amazon in more direct competition with SpaceX's Starlink. So Europe might benefit from that rivalry — more competition usually means better services and lower prices for consumers.

Samantha: The scale of this acquisition tells you how seriously Amazon takes connectivity as a business. Nearly eleven billion dollars is not a side project.

---

Intel is cooking up its next-gen desktop chips — but the kitchen is on fire.

Bob: And finally, something for the hardware enthusiasts. Intel is preparing its next-generation Core Ultra 400 Series desktop processors, codenamed Nova Lake-S, for the second half of this year.

Samantha: Nova Lake-S. These codenames keep getting more dramatic. What do we know about the specs?

Bob: Internal details point to twelve Xe3P cores, which is Intel's latest graphics and compute architecture. It's an ambitious configuration aimed at delivering a significant jump in both performance and efficiency. But here's the twist — this launch is happening against the backdrop of a pretty brutal hardware market.

Samantha: Right. There's an ongoing DRAM crisis, general supply chain issues, and inflationary pressures hitting PC components hard. So Intel is launching a flagship desktop chip into a market where customers are already paying more for everything else in their system.

Bob: It's tricky timing. On the other hand, you could argue Intel needs this launch precisely because the market is tough. They need to give people a compelling reason to upgrade despite higher costs.

Samantha: For European consumers and businesses, the energy efficiency angle is actually quite relevant. Electricity prices in Europe are still elevated compared to historical norms, and the EU's energy regulations keep getting stricter. A chip that delivers more performance per watt isn't just a nice spec sheet number — it's real money saved on energy bills.

Bob: European system integrators and PC manufacturers will be watching closely too. Companies like Fujitsu's European operations or smaller boutique builders need competitive Intel silicon to offer alternatives to AMD. Healthy competition in this space benefits everyone.

Samantha: I'll say this — there's something almost charming about a good old-fashioned CPU launch in a news cycle dominated by quantum AI and personal superintelligence. Sometimes you just need a fast desktop processor.

Bob: Nothing wrong with a chip that makes your spreadsheets fly, Samantha.

Samantha: Spoken like a man who loves his spreadsheets.

---

Samantha: Today we covered: NVIDIA's open-source quantum AI breakthrough, Broadcom and Meta's multi-gigawatt AI chip partnership, Amazon's AI-powered antibody design platform, the Lucid-Uber robotaxi expansion, Amazon's near-eleven-billion-dollar satellite acquisition, and Intel's next-gen desktop processors heading into a stormy market.

Bob: Want to know more or react? Visit stateoftech.eu or email us at info@doorzetters.net.

Bob: State of Tech — the tech world in 15 minutes.