Nvidia’s equity and supply-chain investment commitments have surpassed $40 billion over the course of 2026, a staggering sum that reflects CEO Jensen Huang’s conviction that the AI boom is far from over. But a growing chorus of skeptics argues that all this spending might be doing something more subtle and more concerning: artificially inflating the very demand it claims to be chasing.
When the biggest GPU supplier in the world also becomes one of the biggest investors in the companies buying those GPUs, the line between organic demand and self-referential hype gets blurry fast. And for crypto markets, where Nvidia’s hardware underpins everything from Bitcoin mining to AI-adjacent token projects, the distinction matters more than most people realize.
The investment machine
Nvidia’s most eye-catching move came on January 26, when the company acquired an additional $2 billion in CoreWeave shares at $87.20 each, nearly doubling its ownership stake in the AI cloud provider. CoreWeave is one of the largest buyers of Nvidia GPUs on the planet. So Nvidia is essentially funding a company whose primary business is purchasing Nvidia products.
Across 2026, Nvidia’s total investment commitments have blown past $40 billion, spanning equity positions and supply-chain deals designed to lock in the AI infrastructure buildout for years to come.
In May, Nvidia reported earnings that prompted an $80 billion stock buyback authorization alongside a dividend hike. Data center revenue, powered by the ramp-up of Nvidia’s Blackwell chip architecture, pushed quarterly revenue forecasts to roughly $91 billion.
Mid-2026, Huang stated that the AI market is “a long way from” being a bubble and described demand as “incredibly strong.”
The crypto connection runs deeper than you think
In July 2026, miners like Hut 8 and IREN secured new AI data center contracts worth billions. These companies, which built massive GPU and power infrastructure for proof-of-work mining, are now pivoting toward AI workloads that generate far more predictable revenue streams.
This represents a fundamental reallocation of compute resources away from pure crypto mining and toward AI inference and training. For Bitcoin’s hashrate, that could eventually mean reduced competition among miners as some operators chase higher-margin AI contracts instead.
What this means for investors
The bull case is straightforward. AI workloads are genuinely expanding, and Nvidia’s Blackwell chips represent a generational leap in performance-per-watt. Quarterly forecasts approaching $91 billion suggest the revenue is actually materializing, not just being projected.
The bear case centers on hyperscalers — the Amazons and Microsofts and Googles of the world — which have been ramping capital expenditures on AI infrastructure at a pace that some analysts consider unsustainable. And Nvidia’s own investments in downstream customers like CoreWeave make it harder to distinguish between genuine market pull and financially engineered demand.
For crypto-native investors, sustained AI infrastructure buildout is broadly positive for companies like Hut 8 and IREN that are diversifying into AI services while maintaining crypto mining operations. Their ability to monetize existing GPU infrastructure across multiple revenue streams makes them more resilient to any single sector’s downturn.
The $80 billion buyback authorization is worth watching closely. Stock buybacks at this scale can mask decelerating growth by supporting per-share earnings even as absolute growth slows, and it’s the kind of financial engineering that demands scrutiny when paired with aggressive downstream investing.
Nvidia’s revenue is real. The Blackwell chips are real. But so is the circular logic embedded in a company that invests billions in its own customers while simultaneously citing their purchases as evidence of unstoppable demand. Investors across both tech and crypto would be wise to track not just Nvidia’s top-line results, but the composition of that demand — specifically how much of it traces back to Nvidia’s own capital.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.

1 hour ago
6







English (US) ·