Stack Alliances: NVIDIA x Microsoft

Do you remember the Intel Inside campaign of the 1990s?

At the time, Intel realised they were selling a commoditised product – and they had to put themselves in a better position than the competition.

The Intel Inside Program

In short, PC makers that were partners of the Intel Inside program were reimbursed a part of their advertising costs. In exchange they had to mark their computers with the famous Intel Inside sticker.

By doing this, Intel transformed itself from just an invisible Chip maker to a household brand.

Buyers of computers would then start to ask: "Is it Intel Inside?"

Nvidia Inside... (and Outside)

Nvidia just announced an AI-powered series of consumer computers. These computers will be powered with Nvidia's RTX Spark – a System on Chip (SoC) which puts together:

• Blackwell GPU technology
• an ARM-powered CPU
• and unified memory, on a single chip

This chip is purpose-built to run AI agents, frontier models, creative workloads and next-generation applications locally on the device.

In Zooming out on AI: Part #1 we discussed Big-Tech Complex wars – explaining the complicated game theory and competitive dynamics between Big Tech/AI.

The big carve out at the moment is Google vs anti-Google. We have Google and Apple on the one hand, with basically everyone else on the other.

With the recent Microsoft - OpenAI divorce, we see a new configuration emerging in the pre-existing Microsoft x NVDA collaboration.

In the previous phase of the AI cycle, and yes, I think we are entering a new phase: Microsoft purchased infrastructure from NVDA to power its Data Centers, and by extension its baby, OpenAI.

Recently, OpenAI forged a major partnership with AMD to purchase up to 10% of the company if they fulfil their obligations under the OpenAI - AMD deal.

You see, Sam needed a captive Semiconductor company in the OpenAI stack – because he knew Big Tech is more powerful than him, and it would eventually matter. Owning 10% of AMD will be a major strategic unlock for Altman and OpenAI in this great battle...

--> OpenAI still works with both Microsoft and Nvidia – but is diversifying by layering more partners up and down the stack.

Sam Altman's OpenAI will *not* have a single partner holding all the cards...

...Whether that's Jensen's Nvidia or Satya's Microsoft.

In the meanwhile, Jensen understands that the Data Center trade (i.e. Nvidia stuffing DC infrastructure to Neoclouds and Hyperscalers) is getting worn out.

• Enterprises are managing their token budgets and cutting costs
• Competitors are entering the marketplace and stealing share from Nvidia
• The CUDA moat could eventually erode
• Custom Silicon is entering the picture more strongly – Google TPU etc.

How does Jensen keep the growth coming as the AI cycle changes gears and competition heats up – while making sure Nvidia technology keeps its dominance across every layer of the AI tech stack?

Besides, Jensen is no stranger to major bear markets in Nvidia stock – he knows Semis are cyclical.

....And no, this time is not different.

Jensen's Coup

So what does Jensen do?