← The State of Tech RSS Feed

9 september 2026 · 14:08

Apple foldable iPhone Ultra, new CEO & Samsung-Mistral

Apple has unveiled its first foldable iPhone, the iPhone Ultra, at today's Surprise and Shine event, marking the biggest change to the lineup in over a decade and the public debut of new CEO John Ternus. In the same news cycle, Samsung has signed a strategic partnership with French AI lab Mistral to embed large language models into its semiconductor manufacturing, a deal that puts European AI at the centre of global chip production. This episode covers both stories plus ASML's High NA EUV commitments, US quantum funding and two practical items worth your time.

Apple Samsung Mistral AI ASML Quantum Computing Arm

Beluister deze aflevering:

Transcript

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

Bob: And I'm Bob Russell. Today: Apple debuts its first foldable iPhone under a brand-new CEO, Samsung ties up with French AI lab Mistral to reinvent chipmaking, ASML locks in TSMC and Samsung for its next-generation lithography machines, Washington puts three hundred million dollars behind quantum computing, and to close out, two things you can actually use today: Arm's new mobile GPU with built-in neural hardware for smarter phone graphics, and Canada's new national AI literacy programme with free public resources anyone in Europe can dip into. Let's start with Cupertino.

Apple unveils its first foldable iPhone under new CEO John Ternus at the Surprise and Shine event.

Samantha: Apple's annual event today is a genuine milestone. It's the debut of new chief executive John Ternus, who took over from Tim Cook on the first of September, and it's the launch of the company's first ever foldable iPhone. That device is expected to be called the iPhone Ultra, and it marks the biggest change in Apple's phone lineup in more than a decade. Alongside it, Apple is showing off new iPhone 18 models with upgraded Siri and fresh AI features, plus updates to the Watch and AirPods.

Bob: The foldable itself is what analysts have been waiting for. Samsung, Huawei and Honor have been in this market for years, and Apple has held back until it could deliver a hinge and a display that feel premium. Investors are watching closely for something else too: proof that Apple can catch up on artificial intelligence. There's been persistent worry about how Siri stacks up against Google Gemini and OpenAI's ChatGPT, and today's demonstrations are meant to reassure Wall Street that the company has a credible on-device AI story.

Samantha: For European consumers, the foldable will land in a premium market that's already crowded, but Apple's brand pull is unique. A folding iPhone at a very high price point will still shift millions of units, and it puts pressure on Samsung's Galaxy Fold line right at home in the EU. The other angle is regulatory. Apple's AI features have to comply with the Digital Markets Act and the EU AI Act, which means some Siri capabilities may launch later or in a different form on this side of the Atlantic, as we've seen before with Apple Intelligence.

Bob: And there's the leadership question. Ternus is a hardware engineer by background, he ran iPhone and iPad development for years, so his fingerprints are all over this launch. The tone he sets today matters. Cook was the operations genius who scaled Apple into a three trillion dollar company. Ternus has to prove Apple can still surprise people, and a foldable is a much bolder statement than another camera upgrade. The market response over the next few days will tell us whether investors buy the story.

Samantha: European carriers and retailers will also be watching pricing carefully. If the Ultra lands well above two thousand euros, that's a signal about where the premium smartphone market is heading overall, and it gives room for European resellers, refurbishers and trade-in programmes to build new bundles around it.

Samsung Electronics partners with French AI startup Mistral to embed AI across its chip manufacturing.

Bob: Samsung Electronics has signed a strategic partnership with Mistral AI, the French startup that has become Europe's flagship AI lab. The deal was unveiled during a state summit in Paris and it puts Mistral's models, including the large flagship model, into Samsung's semiconductor engineering and manufacturing operations. These will be custom on-premises AI systems, meaning the models run inside Samsung's own facilities rather than in a public cloud. Samsung is also making a lead investment in Mistral's latest funding round.

Samantha: The industrial logic is straightforward. Chip fabrication is one of the most complex manufacturing processes on earth, and yield stability, spotting tiny defects and optimising equipment, is where fortunes are made and lost. Samsung wants Mistral's AI to speed up development cycles, catch flaws earlier, and squeeze more good chips out of each wafer. Keeping the models on-site matters because the manufacturing data involved is some of the most sensitive intellectual property Samsung owns.

Bob: For Europe, this is a strong endorsement of Mistral's approach. The startup has positioned itself as the credible European alternative to OpenAI and Anthropic, and landing a customer the size of Samsung, plus a lead investment, validates that pitch. It also helps address the perennial concern that Europe funds AI research but loses the commercial upside to American firms. Here you have a Korean giant paying a French company to solve a real industrial problem.

Samantha: The broader signal is that AI is moving decisively into heavy industry. We've spent a lot of time talking about chatbots and coding assistants, but the biggest economic gains may come from applying AI to physical manufacturing, where a one percent yield improvement can translate into hundreds of millions of dollars. European industrial groups from Airbus to Siemens are watching this kind of deal to figure out their own playbook.

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.

ASML locks in TSMC and Samsung as customers for its next-generation High NA EUV lithography machines.

Bob: Staying with chips, ASML has secured firm commitments from TSMC and Samsung to adopt its latest High NA extreme ultraviolet lithography machines. These are the tools that print the finest patterns onto silicon, and each machine costs around four hundred million dollars. Samsung plans to bring them into high-volume production by 2028, TSMC by 2030. ASML and TSMC also announced they'll work together on a shift to larger, twelve-inch photomasks, which should improve productivity in the fabs and eventually lower the cost per chip.

Samantha: The demand driver here is unambiguous: artificial intelligence. Training and running the largest AI models requires ever more powerful accelerators, and those accelerators need ever finer manufacturing. High NA EUV is currently the only path to keep shrinking transistors at the leading edge, and ASML is the only company in the world that makes these machines. That gives the Netherlands a genuinely strategic position in the global chip supply chain.

Bob: For Europe, ASML remains one of the crown jewels. Any story about the AI boom eventually comes back to who supplies the picks and shovels, and increasingly the answer is a Dutch company based in Veldhoven. The commitments from TSMC and Samsung also matter because they lock in demand for equipment that takes years to build and install, giving ASML predictable revenue well into the next decade.

Samantha: There's a geopolitical layer too. Export controls have restricted which High NA machines can go to which countries, and the Dutch government has been navigating pressure from Washington while trying to protect its own industrial base. Deals like these keep the top customers in Taiwan and South Korea firmly inside ASML's orbit, and they underline how much of the AI revolution depends on tools designed and built inside the European Union.

The US Department of Commerce awards three hundred million dollars in CHIPS Act funding to accelerate quantum computing.

Bob: Time for something outside AI. The US Department of Commerce has finalised agreements to hand one hundred million dollars each to three quantum computing companies: D-Wave, Rigetti Computing, and Quantinuum. That's three hundred million dollars in total, allocated under the CHIPS Research and Development Office. The money is targeted specifically at superconducting quantum computing, the branch of the field that uses supercooled circuits to build qubits. Rigetti, for example, will use its share to shrink readout electronics, expand cryogenic capacity, and design chips with more connections between qubits.

Samantha: These are the deeply unglamorous engineering problems that stand between today's noisy quantum prototypes and something that can genuinely outperform classical computers on real work. Scaling from a few hundred qubits to tens of thousands requires solving all of them at once. Washington is basically saying it doesn't want to leave this to venture capital alone, and it wants to build a domestic technological advantage while quantum is still moving from lab to industry.

Bob: For Europe, the immediate effect is competitive pressure. The European Union has its own Quantum Flagship programme and national initiatives in Germany, France and the Netherlands, and there are strong European players like IQM in Finland and Pasqal in France. But three hundred million dollars, on top of existing US federal support, sharpens the argument that European quantum funding needs to scale up if the continent wants a seat at the table when this technology matures.

Samantha: What makes today's news notable is that this is not seed money for research papers, it's directed at making real hardware more manufacturable. Compared with the hundreds of billions of dollars flowing into AI infrastructure, three hundred million looks small, but in the quantum world it's a meaningful cheque, and it targets exactly the engineering bottlenecks the industry keeps flagging.

Arm builds dedicated AI hardware directly into its new Mali G2 Ultra NX mobile GPU.

Bob: To close out, two things you can actually use in the near term. First, Arm has unveiled a new mobile graphics design called the Mali G2 Ultra NX. For the first time, it puts dedicated AI hardware directly inside the shader cores of the GPU, the part of the chip that draws graphics on your phone. Arm claims up to four times the performance per watt on neural graphics tasks, which is engineer-speak for smoother visuals with less battery drain.

Samantha: What does that mean in practice for your next Android phone? Games can be rendered at a lower internal resolution, then reconstructed by a small AI model into a sharper image with a higher frame rate. It also unlocks better on-device AI features, from smarter cameras to real-time translation, without shipping your data to the cloud. MediaTek is expected to be the first chipmaker to adopt the design, so we'll likely see it inside mid-range and high-end Android handsets over the next year.

Bob: For European users this matters on two levels. Practically, it means Android phones will start closing the gap with Apple on on-device AI, which is good for choice and good for prices. Politically, this technology is designed by Arm, a UK-based company that ends up inside almost every smartphone on the planet. That reinforces Britain's role in the AI hardware story, even after it left the European Union.

Samantha: The Mali G2 Ultra NX itself isn't something you'll download, obviously, but it's worth knowing the name. When you're comparing phones next year and you see the Mali G2 Ultra NX badge or MediaTek's marketing around it, that's the signal that the device can handle serious on-device AI without melting the battery. For anyone playing graphics-heavy games or running local AI tools, it's a genuine capability jump rather than the usual marketing bump.

Canada launches its National AI Literacy Initiative under the AI for All strategy.

Bob: Second useful item today: Canada is launching its National AI Literacy Initiative, led by the country's Minister of Artificial Intelligence, Evan Solomon. It's part of the broader AI for All national strategy, and it's aimed squarely at ordinary citizens rather than researchers. The programme will roll out educational resources across Canada, with the launch event happening in Edmonton. The goal is to help people understand what AI can and cannot do, and how to use it responsibly at work, at school and at home.

Samantha: The reason to flag this outside Canada is that a lot of the material will be published online and free to access. If you're in Europe and looking for plain-language explainers about how large language models work, what bias in an AI system actually looks like, or how to spot AI-generated content, the Government of Canada's resources are typically well-produced, in English and French, and free of any commercial pitch. That makes them useful for parents, teachers, small business owners and anyone who wants a trustworthy starting point.

Bob: For European policymakers, this Canadian initiative is also a useful reference model. The EU AI Act requires member states to promote AI literacy, but leaves the details largely to national governments. Canada's programme offers a working example of how to structure a nationwide campaign: define the audience, produce accessible material, and tie it to a broader national strategy. Expect to see European ministries borrowing ideas from it over the coming year.

Samantha: The point is that AI is no longer a niche technology, and understanding it is becoming a basic civic skill, a bit like knowing how to spot a phishing email. Whether you're a curious teenager or a manager trying to decide whether to let staff use ChatGPT, having a calm, government-backed explainer at hand is genuinely useful, and today's launch makes one more of those available.

Samantha: Today we covered: Apple's foldable iPhone Ultra and new CEO John Ternus, Samsung's chipmaking partnership with Mistral AI, ASML locking in TSMC and Samsung for High NA lithography, three hundred million dollars in US quantum funding for D-Wave, Rigetti and Quantinuum, and two things worth exploring: Arm's Mali G2 Ultra NX mobile GPU with built-in neural hardware, and Canada's National AI Literacy Initiative with its free public resources.

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.