Nvidia Hires Microsoft’s Parker to Steer Global Operations as AI Infrastructure Pivots to Power and Finance

Nvidia has poached Nicholas Parker from Microsoft to run its worldwide operations, handing him a guaranteed compensation package north of $40 million. The hire is the clearest signal yet that the chipmaker no longer sees itself as a mere component supplier. CEO Jensen Huang wants to dominate the entire stack of artificial intelligence infrastructure — from the silicon to the software, and increasingly, the financing and delivery model that gets those systems into customers’ hands.

The move arrives as the stock takes a pause that looks less like a sell-off and more like a collective recalibration. Between early June and early July, the shares shed 12.6%. They closed Friday at €171.98, up 1.09% on the session, but still 15.07% below the all-time high of €202.50 reached in mid-May. The annualised 30‑day volatility of 38.25% reflects a market that is still swinging hard in both directions.

Technically, the stock is caught in a tug‑of‑war. It trades 5.17% beneath its 50‑day moving average of €181.36, yet remains 4.73% above the 200‑day average of €164.21. The relative strength index stands at 43.8, a neutral reading that suggests neither fatigue nor panic. From the 52‑week low of €134.06 set almost exactly a year ago, the recovery has been substantial — though the year‑to‑date gain of 6.75% looks anaemic compared with the explosive returns of earlier cycles.

Behind the price action, the nature of the bottleneck that constrains Nvidia’s growth is shifting. The 2023‑2024 narrative was all about chip scarcity. Today the limit is power — and its cost. Nvidia’s upcoming Rubin GPU and rack system, slated to launch this year, draws roughly 300 kilowatts per rack. Future architectures are expected to push toward one megawatt, enough electricity to power 750 average U.S. homes. The company openly acknowledges that about 30% of the energy entering its data centres never contributes to AI computation; it is lost to cooling and transmission across sprawling campuses.

Should investors sell immediately? Or is it worth buying Nvidia?

Nvidia is engineering its way around the heat problem. The Rubin generation uses full liquid cooling that operates at up to 45 degrees Celsius, eliminating traditional fans and cutting water consumption sharply. But the demand side of the equation is hardly the only challenge. The company is also adapting its sales engine to ease the financial bottleneck that prevents many AI startups from accessing its hardware. It is rolling out a model that lets cloud providers obtain GPUs through revenue‑sharing agreements and credit structures, rather than paying the full cost upfront.

That shift from pure chip sales to infrastructure finance is a quiet but pivotal change. It turns Nvidia into a lender and partner in the buildout, not just a vendor. To reinforce its position further, the company has acquired SchedMD and the Slurm workload manager, adding software tools that lock customers into its ecosystem. The “AI factory” concept — exemplified by a partnership with Naver Cloud in South Korea, where the two are building a digital twin of Seoul — shows how Nvidia wants to sell not just silicon but the blueprint and machinery for entire localised AI economies.

Not everything is moving in lockstep. The Kyber architecture has been pushed back to 2028, underscoring the immense manufacturing complexity involved. Geopolitical headwinds persist: Nvidia can still export H200 chips to China under certain revenue‑sharing arrangements, but it faces increasingly stiff competition from domestic rival Huawei. The long‑term dominance of the platform will depend on how quickly the power grid, the financing pipeline, and the software moat can all be reinforced simultaneously.

Against that landscape, the average analyst price target of €263.59 implies roughly 53% upside from the current share price. Meanwhile, the quarterly dividend of $0.25 per share — with the ex‑date set for June 4, 2026 — provides a small but symbolic reminder that Nvidia is no longer a pure growth story. The stock’s next big move will likely be determined less by how many Rubin or Blackwell units ship and more by how fast the world can supply the megawatts and the money to plug them in.