You Could Buy Nvidia for Its 106% Revenue Growth and 75% Gross Margin. But There's an Even Better Reason the AI Stock Has Room to Run.

Despite sky-high expectations, Nvidia (NASDAQ: NVDA) delivered yet another blowout quarter that featured $96.2 billion in revenue -- up 106% year over year and 18% quarter over quarter, along with a 75% gross margin. The gross margin was especially impressive, considering surging prices for memory chips -- which Nvidia buys and integrates into its rack-scale Vera Rubin platform. However, Nvidia does expect gross margin to tick down to 74% in the upcoming third quarter of fiscal 2027, with revenue at $108 billion, up 12.3% quarter over quarter.

Still, the results were impeccable, and so was Nvidia's guidance for a 70% increase in fiscal 2028 revenue compared with fiscal 2027, driven by surging demand from hyperscalers, artificial intelligence (AI) labs, AI natives, enterprises, and sovereign customers. The guidance reinforces the need for rapidly expanding AI infrastructure and follows up on Nvidia's recently announced partnerships with financial institutions for $500 billion in AI capital investment to build Nvidia's computing and full-stack AI infrastructure. The computing will be sold to AI labs, AI start-ups, AI clouds, and other enterprises that need computing power.

There was a lot to like about Nvidia's report and earnings call. But there's one metric that stood out above the rest. Here's my biggest takeaway from Nvidia's Q2 earnings, and why it cements Nvidia as one of the best growth stocks to buy now.

Image source: Getty Images.

Vera Rubin extends far beyond GPUs

In May, Nvidia announced that its Vera Rubin platform had ramped up into full production, with shipments beginning this fall, aligning with Nvidia's upcoming Q3 fiscal 2027. On the Aug. 26 Q2 fiscal 2027 earnings call, Nvidia confirmed that shipments began in August and expects Vera Rubin to account for 20% of its data center revenue in Q3. For context, data center revenue was 92.5% of Nvidia's total Q2 revenue.

The speed at which Vera Rubin will affect Nvidia's top line was the best part of Nvidia's latest print -- more important than its quarterly results or next quarter's guidance. Rubin represents the fastest product ramp-up in Nvidia's history. And the fact that it's already contributing so much to Nvidia's results shows that demand continues to far outpace Nvidia's supply. So with Rubin sales pouring in for the back half of fiscal 2027 and fiscal 2028, it's unsurprising that Nvidia's guidance came in well ahead of expectations.

While Rubin will undoubtedly have a major impact on Nvidia's near-term results, there's an even bigger takeaway for long-term investors: Nvidia's customers are willing to pay a premium for Rubin, meaning its benefits clearly outweigh the high price tag.

Unlike earlier architectures, Rubin comprises more than just graphics processing units (GPUs). It includes a rack-scale offering for data centers that includes six Nvidia chips -- GPUs, central processing units (CPUs), and interconnects. Rubin is tailor-made for large-scale, cost-effective AI training and inference needed from AI factories. And because Rubin is essentially bundled as a rack-scale plug-and-play offering for AI data centers, Nvidia is capturing a larger share of data center revenue than in the past.

On the Q2 fiscal 2027 earnings call, Nvidia's CFO Colette Kress went into detail about why Rubin is more profitable than its previous platforms:

Our second unique capability is our full-stack AI factory platform that is expanding our share of the data center TAM [total addressable market]. Since Hopper, our revenue opportunity has grown from roughly $18 billion per gigawatt to $25 billion with Blackwell, to $40 billion with Vera Rubin, which now spans Vera CPU, Rubin GPU, NVLink, InfiniBand or Ethernet, and Groq LPU, announced earlier this week. Our ability to extreme co-design across GPU, CPU, NVLink scale-up networking, scale-out networking, systems, algorithms, and software enables us to deliver X factor performance gain every generation. Vera Rubin exemplifies this, delivering 30x higher throughput per megawatt and 35x lower token cost relative to Grace Blackwell Ultra. We commenced production shipments of Vera Rubin earlier this month. Having already received purchase orders from every major hyperscaler, AI cloud, and system OEM [original equipment manufacturer], we expect Vera Rubin to mark the fastest product ramp in Nvidia's history.

Despite ongoing fears that Nvidia's growth would eventually slow, the company continues to prove that its large size is not yet a limiting factor, as its business is evolving from cyclical hardware sales to being the primary provider of AI computing infrastructure.

On the earnings call, Nvidia CEO Jensen Huang said he expects the vast majority of the world's data centers to be Vera Rubin NVL72 -- the official name of the rack-scale supercomputer that consists of 72 Rubin GPUs, 36 Vera CPUs, memory chips, and networking.

Nvidia is well-positioned to capture sales from new AI factories, as well as swap out racks of older Nvidia tech with this latest platform -- once again proving that Vera Rubin really is a monumental breakthrough in AI training and inference rather than a marginal upgrade over Grace Blackwell Ultra.

Nvidia remains a high-conviction buy

While it's easy to get enamored with Nvidia's quarterly results, long-term investors should focus on Nvidia's development pipeline. Nvidia used to rely mainly on one-off GPU hardware sales. Now, Nvidia is expanding into a product and service ecosystem that includes software, AI networking, and other hardware to support AI infrastructure at scale.

Nvidia's innovation shows no signs of slowing, and its business model is now far less cyclical than in the past, supporting high margins and ample free cash flow.

Add it all up, and Nvidia continues to stand out as the foundational AI stock for long-term investors to buy and hold for years, if not decades to come.

Daniel Foelber has positions in Nvidia. The Motley Fool has positions in and recommends Nvidia. The Motley Fool has a disclosure policy.