Trust holding Nvidia released very strong FY3Q22 earnings results after the close on Wednesday, with record sales of $7.103 billion (+50% and breezing past expectations of $6.8 billion) and adjusted earnings of $1.17 per share (+60% YoY) outpacing the $1.10 per share consensus. Additionally, gross margin of 67% (+150bos YoY) was right in line with what the street was looking for.

“The third quarter was outstanding, with record revenue,” commented CEO Jensen Huang on the release. “Demand for NVIDIA AI is surging, driven by hyperscale and cloud scale-out, and broadening adoption by more than 25,000 companies. NVIDIA RTX has reinvented computer graphics with ray tracing and AI, and is the ideal upgrade for the large, growing market of gamers and creators, as well as designers and professionals building home workstations.

Before getting into the results, let’s take a quick look at the guidance because while the headline numbers were pretty great, they almost pale in comparison to the guidance management provided. On the top line, management expects to generate revenue of $7.4 billion +/- 2%, that is ~$600 million ahead of the $6.89 billion expected coming into the print. Additionally, gross margin guidance of 67% was ahead of the 66.4% consensus. On the other side of the income statement, adjusted operating expenses are expected to be $2.02 billion while other expenses are expected to come in at ~$60 million, and capital expenditures are expected to be approximately $250 million to $275 million, slightly above the $213 million consensus, however, more than acceptable given the revenue and gross margin expectations.

Jumping into the results, by segment, Graphics revenue came in at $4.092 billion while Compute and Networking revenue came in at $3.011 billion. That said, Wall Street prefers to group revenues by platform.

Starting with Gaming, record revenue of $3.221 billion (+42% YoY) outpaced expectations of $3.138 billion.
- Sales were driven by GeForce GPUs as Nvidia saw “strong demand for our NVIDIA Ampere architecture products leading into the holiday season.” And before the bears come out and tell you that the demand is coming from crypto miners, note that management highlighted on the release that nearly all of the desktop Ampere-based GPU shipments were “Lite Hash Rate” — meaning that the chips’ potential hash rates are capped in order to disincentivize purchases for mining purposes and direct GeForce cards to gamers. Of course while some crypto mining demand no doubt seeps into the segment, management previously released a dedicated Cryptocurrency Mining Processor (CMP) to mitigate the impact. Those results are housed in OEM & Other ( see below).
- Aiding demand is the fact that more games are coming out that support RTX technology — so, if you want the best graphics possible and you don’t have an RTX capable chip, you’re forced to upgrade, which is what we are seeing. On the call, CFO Colette Kress commented, “NVIDIA RTX technology is driving our biggest ever refresh cycle with gamers, and continues to expand our base with creators. RTX introduced groundbreaking real-time ray tracing and AI-enabled super resolution capabilities, which are getting adopted at an accelerating pace. More than 200 games and applications now support NVIDIA RTX, including 125 with NVIDIA DLSS. This quarter alone, 45 new games shipped with DLSS. And NVIDIA Reflex Latency Reducing Technology is in top esport titles, including 'Valorant,' 'Fortnite,' 'Apex Legends' and 'Overwatch.'”
- In terms of where we are in the refresh cycle, we appear to have plenty of room left to run as Kress noted “We estimate that a 1/4 of our installed base has adopted RTX GPUs. Looking ahead, we expect continued upgrades as well as growth from NVIDIA GeForce users, given rapidly expanding RTX support and the growing popularity of gaming, esports, content creation and streaming.”
Record Data Center revenue of $2.936 billion (+55% YoY) were also a record and outpaced the $2.738 billion consensus.
- Driving the segment were “sales of NVIDIA Ampere architecture products to hyperscale customers for cloud computing and workloads such as natural language processing and deep recommender models, as well as to vertical industries.”
- Regarding inference technology, Kress added on the call that the company “released the latest version of [their] Triton inference server software, enabling compute-intensive inference workloads such as large language models to scale across multiple ... nodes with real-time performance. Over 25,000 companies worldwide use NVIDIA AI inference. A great new example is Microsoft Teams, which has nearly 250 million monthly active users. It uses NVIDIA AI to convert speech to text real time during video calls in 28 languages in a cost-effective way.”
- Perhaps most telling of the future, despite the incredible results we have seen thus far from Nvidia, Huang quantified the remaining growth opportunity, commenting that every single server will be GPU accelerated and that when you look at all the clouds and all the enterprise data centers out there, less than 10% currently are. That kind of gives you a sense of where you are. In terms of the workloads, it's also consistent with that in the sense that a lot of the workloads still only run on CPUs.”
Professional Visualization revenue of $577 million (+144% YoY) came in ahead of the $533 million analysts were expecting.
- Ampere-based products were once again the main driver as demand for desktop and notebook workstation GPUs was driven by enterprise efforts to deploy systems to support hybrid work environments.
- On the call, Kress noted that “Leading verticals, including media and entertainment, health care, public sector and automotive.
- Moreover, Kress added some commentary on Omniverse, noting that “initial market reception to Omniverse has been incredible. Professionals at over 700 companies are evaluating the platform, including BMW, Ericsson, Lockheed Martin and Sony Pictures. More than 70,000 individual creators have downloaded Omniverse since the open beta launch in December.” As for the opportunity, while that is certainly a strong start, Kress reminded investors that there are approximately 40 million 3D designers in the global market.”
Automotive sales were a weak spot as revenue of $135 million (+8% YoY) was short versus the $160 million estimate.
- While the segment missed expectations, as you might be able to tell given the size of the revenue versus the overall company, this isn’t much of a focus for analysts in the near-term as most of the upside is expected to come in the decade ahead as autonomous vehicles, utilizing Nvidia technology, start to hit the road. On the release, the team commented that the annual growth “was due to the ramp of self-driving programs, while the sequential decline was related to automotive makers’ supply constraints.”
OEM & Other sales $234 million (+21% YoY) was well ahead of the $161 million the street was looking for.
- Nvidia’s CMP, which is used to mine Ethereum, is housed in that segment. That product accounted for just under half ($105 million) of the segment’s performance.

With the platform stuff out of the way, let’s get to the really fun stuff — Huang’s vision of the future. In our preview note, we commented that we would on the lookout for more software commentary that can provide deeper insights into the growth of recurring software revenues, which tend to come with expanded profit margins are key to sustaining the company’s recent multiple expansion. As expected, the analysts wasted no time asking about Omniverse and what we heard was nothing short of incredible — and in our view, justifies the multiple expansion we’ve seen in recent weeks.

Most exciting in our view in terms of both the technology and the future revenue potential was Huang’s commentary around “Omniverse Avatar.” On the call, Huang commented that while many people think of robots as physical creations, he believes that the largest application of robot technology will be in the from of avatars, commenting, “we built Omniverse Avatar to make it easy for people to integrate some amazing technology from Consumer Vision speech recognition, natural language understanding, gesture recognition, facial animation synthesis, recommender systems, all of that integrated into one system and running in real in time. That avatar system is essentially a robotic system. And the way that you would do that is, for example, the 25 million-or-so retail stores, restaurants, places like airports and train stations and office buildings and such, where you're going to have intelligent avatars doing a lot of assistance? They might be doing a check out, they might be doing check in, they might be doing giving customer support. And all of that could be done with Avatars.”

Now, while that may sound incredibly ambitious — and it is — what is important to note is that Nvidia has already built most of the technology required to make this a reality in front of our very eyes, and Omniverse Avatars are really just the realization of all the company's existing software stacks. For example, some of the parts of an Omniverse Avatar include the Merlin recommender software, Megatron the large language model, Riva the speech AI, Phase eye tracking technology, the company’s slew of computer vision AI software stacks, the natural speech synthesis technology that we see every year at GTC when “I AM AI” takes us through the opening credits.

What this all amounts to is a potentially massive licensing opportunity for Nvidia with Huang stating that an Omniverse Avatar could cost say $1,000 per year (not a bad salary from an employer’s point of view). Now multiply that cost per avatar per year by say “25 million or so places where you would have a digital Avatar as customer support or checkout, smart retail, smart warehouse, smart whatever it is.” Then add in the fact that you have some 100 million cars, each of which can potentially have something like an Omniverse Avatar. Of course it doesn’t end there with Huang adding that “then, of course, behind all of that, call it a couple of hundred million digital agents, intelligent agents in robots and Avatars, at $1,000 per agent, per year, behind it are NVIDIA GPUs and PCs. NVIDIA GPUs in the cloud and NVIDIA GPUs for Omniverse servers.” Start doing the math and you realize that Jensen is targeting a recurring revenue opportunity north of $100 billion per year! No wonder, Huang called it “one of the largest opportunities we've ever seen.”

Of course, while that sounds incredible, what matters to Wall Street is when we might actually see those revenues materialize. In Huang’s view, it won’t be that far into the future, commenting, that in his view we can see these Omniverse Avatars, which don’t make mistakes, never get tired and are always on “in drive-throughs of restaurants, fast food restaurants, checkouts of restaurants in retail stores all over the world within less than 5 years.”

Now consider that potential revenue stream along with the ~90% of servers that have yet to implement GPU acceleration technology, growing adoption of gaming, rapid growth we are seeing on the pro visualization front and the fact that the automotive segment is still in its infancy, and compare that upside to the ~$26 billion analysts expect Nvidia to generate this year and you start to realize that as incredible as the growth story at Nvidia has been to date, we are still in the very early innings.

Bottom line, while we wouldn’t be surprised to see shares consolidate around current levels as better-than-expected headline numbers with strong guidance are what investors no doubt anticipated as they bid shares up more than 30% over the past month. This was an incredible quarter that should serve to solidify the gains. The outlook is as bright as ever and we expect further upside in the quarters ahead as analysts start to revise their models to factor in the recurring revenue upside in future years and attempt to quantify the addressable market Huang has laid out with his vision for the Omniverse.

The CNBC Investing Club is now the official home to my Charitable Trust. It’s the place where you can see every move we make for the portfolio and get my market insight before anyone else. The Charitable Trust and my writings are no longer affiliated with Action Alerts Plus in any way.

As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Typically, Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust’s portfolio. If the trade alert is sent pre-market, Jim waits 5 minutes after the market opens before executing the trade. If the trade alert is issued with less than 45 minutes in the trading day, Jim executes the trade 5 minutes before the market closes. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. See here for the investing disclaimer.

(Jim Cramer's Charitable Trust is long NVDA.)