Nvidia (NVDA -0.85%) stock has rocketed 35% year to date, and that momentum could continue in the coming months. Shareholders just received good news from President Donald Trump regarding export restrictions. They also got good news from Wall Street about how much money hyperscale cloud companies are spending on artificial intelligence (AI) infrastructure.
Here's what investors should know.
Image source: Official White House photo by Joyce N. Boghosian.
President Trump will let Nvidia sell more chips in China
Since 2022, the U.S. government has progressively tightened export controls on AI technologies to keep sophisticated semiconductors and equipment away from the Chinese government and military. For Nvidia, one of the most jarring blows came earlier this year when the Trump administration banned the sale of its H20 GPUs in China.
Nvidia took a $4.5 billion charge in the first quarter associated with excess inventory, and CFO Collette Kress warned, "Losing access to the China AI accelerator market, which we believe will grow to nearly $50 billion, would have a material adverse impact on our business going forward and benefit our foreign competitors worldwide."
However, President Trump recently reached an agreement with CEO Jensen Huang that will let Nvidia sell H20 GPUs in China in exchange for 15% of the revenue generated from the sales going to the U.S. government. Put differently, Nvidia will pay the U.S. government for access to China. Of course, that raises an important question: What happened to the national security issues that made export restrictions necessary in the first place? Regardless, this development is still positive for Nvidia shareholders.
Better yet, President Trump is reportedly open to a similar arrangement with Nvidia Blackwell GPUs. The company would have to build a scaled-back version of the chip, just as the H20 GPU is a scaled-back chip based on the Hopper architecture. But the Trump administration seems willing to let Nvidia operate in China for the foreseeable future, and that's good news for shareholders because China is the second-largest AI market in the world.
Wall Street analysts anticipate more heavy spending on AI infrastructure
Hyperscalers are companies with enormous data center footprints. Amazon, Microsoft, and Alphabet are the three most prominent, but the list also includes Meta Platforms, Oracle, CoreWeave, Apple, and several other companies that provide cloud computing services on a large scale.
Investors closely track hyperscalers' capital expenditures because they have large budgets and their spending decisions provide insight into future trends. With June-quarter financial results now finalized, capital spending among the 11 largest hyperscalers is on track to reach $445 billion this year, up 56% from last year, according to Morgan Stanley.
That is encouraging news for Nvidia shareholders. Before the June quarter, capital spending among the 11 largest hyperscalers was projected to increase 44% to $410 billion in 2025. In other words, the companies with the largest data center footprints are spending more than Wall Street expected, and most of that money is going toward AI infrastructure.
Looking ahead, Bank of America estimates spending on data center AI systems will grow at 26% annually through 2030, with nearly 90% of the total sum going toward AI accelerators and networking hardware. Nvidia has a leadership position in both markets. It accounts for about 80% of AI accelerator sales and over 50% of generative AI networking equipment sales. So the company is well-positioned for future sales growth.
Here's the bottom line: Nvidia may have opportunities in China that the market previously discounted, and capital spending among hyperscalers is tracking ahead of what Wall Street anticipated. As a result, several analysts have raised their earnings forecasts for Nvidia. The consensus estimate now says its adjusted earnings will grow at 43% annually through the fiscal year ending in January 2027. That makes the current valuation of 57 times adjusted earnings look reasonable.
