Nvidia’s Next Leg of Growth Could Be Massive

Over the past three years, demand for Nvidia’s GPUs has ballooned so fast that the company has entered a rare feedback loop. Every architecture upgrade sparks a new wave of orders from hyperscalers, which funds even faster improvement, which attracts even more demand. How long can this last?

If Jensen Huang is right, it isn’t slowing. In fact, Nvidia may be on the verge of its most aggressive revenue surge yet.

Key Points

• Nvidia has roughly $307 billion in upcoming data center revenue tied to Blackwell, Rubin, and its high-margin networking products.

• The company’s data center business has scaled at a historic pace.

• Nvidia may still be undervalued relative to its earnings trajectory.

The Surprise That Shocked Wall Street

Nvidia’s pipeline for its latest chips sits at roughly $500 billion. And not over a decade, over the next five quarters.

Investors barely had time to process the number before the stock jumped and Nvidia crossed the $5 trillion valuation mark. For perspective, that pipeline alone is larger than the entire U.S. semiconductor market in 2019.

But, as usual, the details matter.

A Closer Look Behind the Headline

About 30% of that “order book” includes chips already shipped, meaning Nvidia has recognized some of the revenue already.

Another chunk is tied not to GPUs like Blackwell and Rubin, but to Nvidia’s lesser-known networking products. These include InfiniBand, the ultra-fast data highway that connects GPUs inside AI clusters.

These networking systems are becoming one of Nvidia’s highest-margin businesses. In fact, InfiniBand is so indispensable that several hyperscalers estimate that switching away from it could delay training runs by weeks or even months.

After netting out what’s already recognized, analysts suggest the truer backlog sits at over $300 billion, still an enormous figure, and still expected to be recognized within roughly a year.

3 years ago, Nvidia generated less than one tenth of that annually across the entire company. Today, its data center arm alone clears that every quarter.

That’s not normal business growth, that’s a structural shift in global computing.

What Most Investors Are Missing

Investors are rightly cautious about bold numbers tossed out at conferences. But the more important piece isn’t the exact dollar value, it’s the direction of demand.

There’s increasing evidence that analysts have underestimated how quickly AI infrastructure is scaling. Cloud providers now talk in 12-month cycles instead of 3- to 4-year architectural refreshes. Meta, Google, Amazon, Microsoft, all are projecting AI compute needs that rise faster than power grids can accommodate.

Some data center operators have started publicly discussing electricity procurement as their new bottleneck. Nvidia’s GPUs are no longer the constraint, power availability is. That’s how fast AI demand is growing.

This dynamic works in Nvidia’s favor. Whenever energy capacity opens up, the immediate next step is ordering another batch of accelerators, and Nvidia remains the default vendor.

Given that context, a forward P/E of 30 looks oddly plain for a company that keeps beating its own high bars and may still be undervalued relative to its earnings trajectory.

So, Is Nvidia Still a Buy?

Long-term investors should think less about whether Nvidia hits as much as $500 billion in revenue over the next year, and more about what that backlog says about AI demand.

Nvidia has gone from a mid-sized GPU supplier to the backbone of the world’s new computing standard in about three years. It has become the Wintel of the AI era, except with a monopoly on the accelerator stack, the networking layer, and the software ecosystem (CUDA) that developers build on.

If the Street’s current expectations prove too conservative, and Huang’s comments suggest they might be, Nvidia could still have years of outsized growth ahead.

From where things stand today, Nvidia remains one of the strongest buy-and-hold candidates in the tech sector. For investors building a long-term portfolio around structural megatrends, it deserves serious consideration as a core position.