Meta and Nvidia Pivot to Compute-as-Service to Monetize AI Spend

Meta Platforms saw its stock price climb 9% following the announcement of a new strategic initiative to sell its excess AI compute capacity through a specialized cloud push. After informing investors in April of plans to spend up to $145 billion on infrastructure, the company is seeking to monetize its massive GPU clusters by offering power to external developers. This move signals a shift in Meta’s strategy, evolving from a pure social media and metaverse company into a foundational infrastructure provider for the broader AI ecosystem.

In a similar move to capture the startup market, Nvidia has unveiled a program that allows cash-strapped startups to swap a slice of their future revenue for immediate access to Nvidia’s highly coveted compute power. This "revenue sharing" model aims to lock in the next generation of AI firms into Nvidia’s ecosystem while providing the hardware giant with potentially lucrative equity-like stakes in emerging technologies. Both Meta and Nvidia are increasingly acting as "AI landlords," controlling the essential resources required for model training and deployment.

Meanwhile, legacy software giant SAP is moving in a different direction, embarking on a major restructuring program to rein in traditional operational costs and refocus its capital on AI investments. The restructuring, which has affected thousands of jobs, is part of a broader corporate trend where mature tech companies are cannibalizing their established divisions to fund the high-stakes arms race in generative AI. These strategic shifts across Meta, Nvidia, and SAP underscore a new era of corporate resource allocation where compute power is treated as the primary currency of growth.