The Cheques Are for the Land Not the Intelligence

AI Datum Point

SubscribeSign in

The Cheques Are for the Land Not the Intelligence

AI Datum Point<br>Jun 28, 2026

Share

In October 2026, Google begins paying SpaceX $920 million per month. The contract, disclosed in a SpaceX SEC filing, runs through June 2029 and buys access to approximately 110,000 Nvidia GPUs, CPUs, memory, and related components. At that rate, the full contract is worth roughly $30 billion. What Google is buying is not a model, not a product, not a research programme. It is hardware capacity, secured years in advance, at a price that eliminates flexibility. That is not how you invest in a technology. It is how you acquire a scarce physical asset before someone else does.<br>The same logic runs through the India commitments. Microsoft has pledged $17.5 billion over four years. Google has pledged $15 billion. Amazon has committed over $35 billion through 2030. The combined figure is $67.5 billion, directed at a single geography. India's AI ambitions provide the political cover. The hardware constraint is the actual driver.<br>The energy wall is where this becomes a structural condition rather than a strategic preference. AI data centre expansion at scale runs into grid capacity limits that capital alone cannot resolve. The conflict between hyperscalers and grid operators over power allocation is not a negotiation in progress. It is a physical constraint that multiple independent operators have hit separately, from different directions, and are now routing around through the same set of responses: off-grid generation, behind-the-meter compute, co-located power and cooling systems designed to operate outside grid dependency. Each workaround is a separate actor arriving at the same conclusion. The grid cannot carry this load.<br>The public narrative around AI investment treats the current moment as a capability race. The stated positions of the actors writing the largest cheques do not support that framing. A capability race would look like competing research programmes, differentiated model architectures, proprietary training data at scale. What the record shows is a coordinated multi-actor move toward physical infrastructure ownership: compute, power, cooling, and the real estate to put it on. The model on top is the product that justifies the acquisition. But the acquisition is what actually transfers value between the actors who understand what is happening.<br>India's IT sector has named the gap plainly. The profitable layer of AI is not capability. It's deployment: the integration work, the workflow adaptation, the customisation required to translate a model into a business outcome. That work is labour-intensive, geographically distributed, and resistant to the economies of scale that govern hardware. It is also where the actual revenue is. The hyperscalers have the hardware. The deployment layer remains unowned at scale, and India's $300 billion IT industry is moving to take it.<br>Arm claims up to 65% better price-performance and as much as 60% greater energy efficiency for Arm-based cloud instances across AI workloads, figures it has promoted via The Register and its own marketing materials. Those numbers come from Arm itself and have not been independently audited. But the direction they point to is confirmed by the behaviour of the actors spending the money: Google, Microsoft, and AWS are all now deploying custom Arm-based silicon in their own data centres. The hardware the current investment cycle is built on is not the hardware the next phase runs on. The actors locking in multi-year GPU contracts at $920 million per month know this. They are acquiring positional advantage before the stack shifts, not after.<br>The capital story being told about AI is a story about intelligence. The capital flows tell a story about land.

Share

Previous

Discussion about this post<br>CommentsRestacks

TopLatest

No posts

Ready for more?

Subscribe

© 2026 AI Datum Point · Privacy ∙ Terms ∙ Collection notice<br>Start your SubstackGet the app<br>Substack is the home for great culture

This site requires JavaScript to run correctly. Please turn on JavaScript or unblock scripts