It should come as no surprise to my followers that NVIDIA (NVDA) did it again with another beat-and-raise quarter.

Despite the rumblings of an "AI bubble" or questionable vendor financing tactics, the Starship Jensen delivered real and sustainable growth that pushed the stock to a $2 trillion market cap.

Recall that was my forecast last spring...

Nvidia DGX: Workhorse of AI Will Drive NVDA to $2 Trillion

But even I didn't think it was going to happen this soon.

So what's going on?

You can review that video and article to understand the 2023 business mechanics and industry dynamics of GPUs stacked in DGX systems, running "massively parallel architectures" on CUDA.

I made that piece when NVDA shares were trading $400 to explain to investors why the stock launched in late May from $300 -- and would never look down again.

The pivotal moment came in late May at the Computex event in Taiwan when NVIDIA CEO Jensen Huang revealed that Microsoft, Alphabet, and Meta Platforms were all lined up for his latest DGX-GPU system, the GH200, a large-memory AI supercomputer that uses NVIDIA NVLink to combine up to 256 NVIDIA Grace Hopper Superchips into a single datacenter-sized GPU.

Since I had just published my special report on ChatGPT the day before NVIDIA's March GPU Tech Conference (GTC), I put two and two together... to get 200!

You see, when I published that report on March 20, I had no idea that Jensen would start talking about ChatGPT as “the AI heard round the world” and that this was AI's "iPhone Moment" of transformational demand.

And even though I didn't need any convincing to buy and hold NVDA shares (we had just scooped it near 18-month lows under $125 in October of 2022), something hit me when I heard that those 3 datacenter giants were the first big customers for GH200 systems...

If they were going to keep building and improving their own internal LLMs (large language models), then the other 997 corporations in the Fortune 1000 would be itching to do it too!

Make that the other 1,997 enterprises in the Global 2000.

From March to June, we saw NVIDIA sales projections for last year and this year essentially double. And I said in June that even those expectations were too tame.

Because the race was on for every enterprise to build and scale their own LLM for internal data mining, industry insights, and next-gen innovation.

What's Different Now?

Believe it or not, not much has changed. The demand is still off the charts and the GPUs keep getting better and faster.

And so that means only the important things have changed: the speed, power, names, and price of the GPUs.

Yep, as they get better, they keep getting pricier. And that's why NVIDIA was able to boast 76% gross margins last week in their quarterly report.

On Monday, I made a video using a recent NVIDIA presentation to highlight the secular drivers of this incredible AI revolution.

In that video, I showed over half-a-dozen slides that really communicate the transformation that is occuring for every sector of the economy.

And while some folks like to make fun of ChatGPT for what it can or can't do, this is about far more than generative AI questions and answers.

What I left out of my 7-minute video was a discussion of all the much more important transformations happening in science, engineering, medicine, agriculture, climate studies, and industrial productivity and automation.

Those are the most exciting areas where the "massively parallel architectures" of GPU-driven AI can make a difference and change the world.

Estimates: Up, Up, and Away

In the video, I also show the Zacks Detailed Earnings Estimates tables for NVDA with the stunning upward revisions that we are witnessing since their report last week.

For this fiscal year 2025 (began in February), the topline moved from $93 billion to $98.6 billion. But I said they would keep going higher, and as of Wednesday we're now looking at a consensus of $102.4B for a potential 68% annual leap, with the high estimate at $107B.

And next year is clocking in at over $117B with the high estimate at $136B. Again, I still expect these estimates to all move higher.

In the video, I explain why NVIDIA will likely hit $200 billion in revenues sometime by 2027.

And that means NVIDIA will be the next $3 trillion company.

Disclosure: I own shares of NVDA for the Zacks TAZR Trader portfolio.

More By This Author: