👀In Case You Missed It

  • OpenAI + Oracle unveil “Stargate” — a 1+ GW AI campus in Michigan
    A joint hyperscale project designed to anchor OpenAI’s long-term compute expansion in the U.S. Midwest, reinforcing Oracle’s return as a top-tier cloud partner.

  • Amazon’s Project Rainier deploys ~500,000 Trainium2 chips
    The in-house silicon push reportedly boosts Anthropic’s compute by nearly 5×, confirming Amazon’s strategy to scale with custom accelerators alongside Nvidia GPUs.

  • Canva launches its own foundation model and makes Affinity free
    The platform now runs a proprietary design-aware AI that produces fully editable layered outputs, adds 3D and spreadsheet tools, and integrates an assistant across every workflow.

Read more here

🧠 Cloud & Infrastructure Watch

  • Big Tech earnings confirm an AI build-out sprint, with Alphabet’s first $100B quarter, Amazon’s strongest cloud growth in three years, and Microsoft revenue up 18%, as firms collectively drive an estimated $400B in AI infrastructure spend this year.
    Read More

  • Cisco highlights new fibre-based scale-up networking to cut latency across GPU clusters as operators re-architect data flow for AI-scale systems.
    Read More

  • OpenAI x AWS multi-year compute pact continues to ripple across infra circles.
    Read More

  • Lambda–Microsoft: multi-billion-dollar agreement to deploy tens of thousands of Nvidia GPUs for AI workloads.
    Read More

💻 Hardware & Semiconductor Moves

  • Starcloud-1 becomes the first satellite with an Nvidia H100 in orbit—spaceborne accelerated compute is now live.
    Read More

  • SpaceX signals intent to build data centers in orbit (“will be doing this”), further normalising space DC roadmaps.
    Read More

  • Microsoft signs a $9.7B capacity deal with Australian firm IREN for Nvidia GB300 GPU infrastructure, to be deployed in phases through 2026.
    Read More

  • Microsoft and Lambda complete a multibillion-dollar AI infrastructure deal involving tens of thousands of Nvidia GB300 chips.
    Read More

  • The U.S. grants Microsoft permission to ship Nvidia’s most advanced AI chips to the UAE, widening global access to Blackwell-class GPUs.
    Read More

  • Eaton acquires Boyd Thermal for $9.5B, expanding its liquid cooling and thermal management portfolio for high-power processors.
    Read More

  • Analysts forecast Nvidia GPU shipments to double in the next 12 months, with some projecting potential valuations up to $8.5 trillion as AI demand accelerates.
    Read More

📰 News of the Day

OpenAI Strikes Record $38 Billion Cloud Deal with AWS

Summary:
OpenAI has signed a seven-year, $38 billion agreement with Amazon Web Services to secure massive compute power for the next generation of ChatGPT, Atlas, and agentic AI workloads. The partnership grants OpenAI access to hundreds of thousands of Nvidia GB200 and GB300 GPUs across AWS data centres, with flexible scaling to meet surging demand. The deal marks OpenAI’s largest diversification away from Microsoft’s Azure platform and signals an intensifying race among hyperscalers to dominate AI infrastructure.

Why It Matters:
This is the biggest cloud contract in history and a defining moment for global AI infrastructure. It cements AWS as a primary compute provider for OpenAI, reshapes the balance of power between the major clouds, and further tightens the global GPU supply chain. For AWS, the contract locks in multi-year, high-margin capacity sales and positions it alongside Microsoft as one of the two essential backbones of OpenAI’s expansion.

The Bigger Picture:
AI infrastructure spending has entered an unprecedented acceleration phase — with Microsoft, Amazon, and Google now committing over $400 billion annually to GPUs, energy, and data centres. As OpenAI prepares for a potential $1 trillion IPO, these long-term compute alliances are laying the groundwork for global agentic AI platforms that could reshape both digital labour and sovereign compute strategies for the next decade.With global cybersecurity shortages, Aardvark could redefine defensive automation for enterprises and governments alike — but it also raises red flags about autonomous offensive potential and control boundaries in agentic systems.

Subscribe to keep reading

This content is free, but you must be subscribed to ELI'S AI DAILY | #1 Newsletter in the Tech World to continue reading.

Already a subscriber?Sign in.Not now

Reply

Avatar

or to participate

Keep Reading

No posts found