🪟 Windows just got its M1 moment at Computex

Good morning, tech builders. Nvidia walked into Computex and did what no one quite expected: it launched a chip designed to run AI agents locally on your laptop. Shares of AMD, Intel, and Qualcomm dropped between 4.9% and 8.5% the same afternoon. HP and Dell each rose more than 7%.

That's the shape of a category disruption, not just a product launch. If Nvidia's RTX Spark proves out, the next enterprise PC refresh cycle won't be a spec comparison between AMD and Intel. It'll be a platform bet on who owns the local agent runtime. So: is this the Windows M1 moment, or is it just another expensive promise arriving at the worst time?

In today's recap:

LATEST DEVELOPMENT

Nvidia unveiled the RTX Spark PC chip at Computex in Taipei, co-developed with MediaTek after three years of collaboration with Microsoft aimed at reinventing the PC for the AI era. The chip is designed to run autonomous AI agents locally rather than relying on the cloud, and it's arriving in laptops and compact desktops from Dell, HP, Lenovo, ASUS, Microsoft Surface, and MSI this fall.

AMD, Intel, and Qualcomm all dropped between 4.9% and 8.5% on the day of the announcement while HP and Dell each rose more than 7%. The market read it as a real shift in who owns the PC silicon stack. If local agent compute becomes the default expectation by 2027, every enterprise refresh cycle from here becomes a platform battle Nvidia wasn't a part of two years ago.

→ The Verge

Most AI newsletters summarize headlines. MavSource is different.

We aggregate updates from all major AI newsletters, podcasts, company news, AI labs, public and private company activity, GitHub projects, funding rounds, earnings calls, and investor letters — hundreds of sources every day. Then we summarize what matters, analyze emerging trends, and add our own founder commentary so you understand why a development may matter — not just what happened.

One 5-minute email, every morning. Built for investors, founders, and operators who want to understand AI as a business, technology, and market trend — not just another news cycle.

The daily email is free. It's also the entry point to a deeper intelligence product covering watchlists, public-company read-throughs, startup trackers, and investor-letter tracking.

Subscribe Now For Free

GitHub Copilot moved to usage-based billing on Monday, ending the era of flat $10-$39 monthly subscriptions. Requests are now priced dynamically based on the model, the complexity of the query, and the size of the context submitted. One developer on the $39-per-month Pro+ plan burned 16% of their monthly credit allowance in a single session with Claude 4.8 to fix two bugs, reporting the tool "gave some pretty mediocre suggestions" before the credits vanished.

GitHub's official reasoning is that agentic workflows consume far more compute than prompt-answering, so flat pricing was never sustainable. That's true, but it shifts the cost unpredictability entirely onto developers, who are now routing to OpenRouter, LM Studio, and RooCode to manage the overage. If Copilot's most engaged power users leave, Microsoft will have traded retention for margin at exactly the moment it needs developer loyalty to defend the AI coding market.

→ The Register

Alphabet launched a three-part $80B equity raise on Monday: a $10B private placement to Berkshire Hathaway ($5B in Class A shares at $351.81, $5B in Class C at $348.20), a $30B concurrent public offering split between convertible preferred stock and common shares, and a $40B at-the-market program launching in Q3. Berkshire more than tripled its Alphabet stake last month, bringing it to $16.6B before this transaction.

When Berkshire buys into AI infrastructure at this scale, it's not speculative positioning, it's a vote that the returns on AI capex are calculable within a normal investment horizon. Alphabet has already raised more than $85B in debt over the last year and is targeting $180-190B in annual capex. Issuing equity instead of more debt signals confidence that Alphabet can grow into the valuation rather than lever up against it.

→ Reuters

The RTX Spark chip debuts in consumer hardware from Dell, HP, Lenovo, ASUS, Microsoft Surface, and MSI this fall, with Acer and GIGABYTE to follow. Jensen Huang framed the Vera CPU (the companion central processor to RTX Spark) as opening a "$200 billion new market" for Nvidia, with early Vera adopters including OpenAI, Anthropic, and SpaceX. Counterpoint Research's co-founder called the moment Nvidia's "iPhone or ChatGPT" equivalent for personal computing.

The OEM lineup is the validation layer. It's not a reference design without partners; it's seven major PC makers committing to fall shipments. If the category performs even close to expectations, Nvidia's presence in every enterprise IT refresh conversation from 2027 onward is structurally guaranteed. Huang dismissed AI concerns about software engineer hiring as "complete nonsense," betting that the productivity multiplier creates more roles than it eliminates.

→ WSJ

The European Commission plans to propose sovereignty requirements for cloud computing in highly critical state tenders, banking, energy, and healthcare, that could exclude Amazon, Microsoft, and Google from such contracts. The draft, part of the Cloud and AI Development Act to be unveiled Wednesday by EU tech chief Henna Virkkunen, introduces mandatory non-price criteria including requirements for EU-developed software and hardware. It also proposes fast-track data centre approval with preferential grid access for facilities using European-made chips.

Amazon, Microsoft, and Google hold more than 60% of the global cloud market and have each launched sovereignty-compliant joint ventures in Europe to stay eligible. If this passes, those workarounds may not be enough for the most sensitive public sector contracts. Combined with SoftBank's €75B commitment to French AI data centres last week, European AI infrastructure is becoming a domestic policy priority that could reshape procurement for a decade.

→ Reuters

QUICK HITS

🧡 Enjoyed this issue?

🤝 Recommend our newsletter or leave a feedback.

Cheers, Jason

Connect on LinkedIn, & Twitter.