A tectonic shift is occurring in the artificial intelligence industry: financial and technological ties between key players are becoming so dense that they are erasing the familiar boundaries between supplier, client, and investor. The clearest confirmation of this is my analysis of the recent agreement between Google and Marvell Technology, which I consider one of the most significant market events in recent months.
Deal Structure: More Than Just a Purchase
Google has secured the right to acquire up to 58.97 million Marvell shares under an expanded partnership to develop AI processors. This involves components for its own TPU accelerators—from inference processors to storage controllers and networking solutions. However, the key detail is the option structure. Only about 1.36 million shares will be unlocked evenly in the first year. The remainder of the warrant, divided into 240 tranches, is directly tied to purchases: each tranche activates after Google orders generate an additional $500 million in revenue for Marvell. This means that to fully exercise the option, sales volume must reach a staggering $120 billion.
The exercise price is set at $206.58 per share, with a term extending to August 2033. If the warrant is fully exercised, Google could become Marvell's fifth-largest shareholder. The market reacted instantly: Marvell shares rose, while competitor Broadcom's stock fell by more than 5%. This, in my view, is a clear signal that investors perceive the deal as strengthening Google's position in the race for independence from Nvidia and as a serious blow to Broadcom's monopoly in certain segments.
Financing Demand: A New Level of Complexity
This move by Google is part of a broader trend. Chipmakers and hyperscalers are no longer just selling technology to each other. They are providing guarantees, options, and jointly raising capital. Consider the recent AMD-OpenAI agreement or Nvidia's $105 billion guarantee for OpenAI's data center in Ohio. In the latter case, Nvidia not only invests $1.5 billion in SB Energy but also becomes the exclusive supplier, effectively financing its own future demand. This is not circular financing in the classical sense, but, as I note in my reports, it creates an unprecedented level of interdependence that investors have yet to fully appreciate.
Expert Perspective
Such structures allow infrastructure to be built faster than through balance sheets alone, but they also amplify systemic risks. The Bank for International Settlements has already warned about the opacity of such deals and the risks of asset rehypothecation. I believe that in the coming years, we will see not only a rise in capital expenditures (total AI spending will already exceed $730 billion this year) but also a fundamental restructuring of corporate finance in the technology sector. Competitive advantage is now determined less by model quality and more by the ability to cheaply finance and operate infrastructure. This is the new reality to which the market has yet to adapt.