Key Takeaways From The Nvidia Q3 Earnings Call

Summary:

• General AI market growth and a sustainable competitive advantage are key areas of interest with Nvidia Corporation.
• Blackwell system overheating issues seem resolved, but uncertainty remains about future supply and performance, affecting near-term revenue.
• Weak guidance for next quarter and potential supply challenges raise concerns, but long-term AI demand and market dominance likely remain strong.
• Despite December’s poor historical performance, I maintain a “hold” rating due to Nvidia’s sustained long-term growth potential and underestimated AI demand.

The most anticipated earnings call of the season remains that of Nvidia Corporation (NASDAQ:NVDA). Its share price was roughly flat in the aftermarket following its Q3 earnings, likely because guidance didn’t overly impress, and a revenue beat had been expected. The call was still quite interesting, and I want to highlight key takeaways, especially related to Nvidia’s competitive advantage as well as the growth trajectory of AI hardware. Furthermore, at ~$3.5 trillion, Nvidia’s market cap is so large, while it is still a volatile high-growth stock, it can influence the entire market. An additional reason why I like to keep tabs on it.

One important issue that I wanted to hear something about was the issues Nvidia has had with the new Blackwell system overheating. I got the impression Nvidia fixed those. I don’t understand the technology at a detailed enough level to be certain, but at Tom’s hardware blog they explained the issue like this:

Nvidia’s Blackwell B100 and B200 GPUs link their two chiplets using TSMC’s CoWoS-L packaging technology, which relies on an RDL interposer equipped with local silicon interconnect (LSI) bridges (to enable data transfer rates of about 10 TB/s). The placement of these bridges is critical. However, a supposed mismatch in the thermal expansion properties between the GPU chiplets, LSI bridges, RDL interposer, and motherboard substrate caused the system to warp and fail, and Nvidia reportedly had to modify the top metal layers and bumps of the GPU silicon to enhance production yields. While the company did not disclose specific details about the fix, it did mention that new masks were required.

They were buying into the idea that it would be fixed with sets going out in 2025. These short-term issues are usually not a problem for a growth stock unless they are persisting and threatening the competitive advantage. Earnings right now are not necessarily that relevant, but continued expected dominance is.

On the call, there was a question about it, and CEO Huang answered as follows (emphasis added):

…And so Blackwell demand is very strong. Our execution is on — is going well. And there’s obviously a lot of engineering that we’re doing across the world. You see now systems that are being stood up by Dell and CoreWeave, I think you saw systems from Oracle stood up. You have systems from Microsoft and they’re about to preview their Grace Blackwell systems. You have systems that are at Google. And so all of these CSPs are racing to be first. The engineering that we do with them is, as you know, rather complicated. And the reason for that is because although we build full stack and full infrastructure, we disaggregate all of the — this AI supercomputer and we integrate it into all of the custom data centers in architectures around the world. That integration process is something we’ve done several generations now. We’re very good at it, but still, there’s still a lot of engineering that happens at this point. But as you see from all of the systems that are being stood up, Blackwell is in great shape. And as we mentioned earlier, the supply and what we’re planning to ship this quarter is greater than our previous estimates…

He said a bit more, but the way I see it, he danced a bit around the answer. He said a lot of positive and interesting things about Nvidia’s products and the future of AI, but I didn’t come away reassured there wouldn’t be Blackwell issues anymore. In addition, the company also said their previous system would continue to sell quite well for a bit. Which can be interpreted as good or that customers need to be more certain about Blackwell. Finally, management guided to earnings in the low 70s until Blackwell really ramps up.

Again, I don’t think it matters much whether the company achieves 71% or 75% gross margins in any given quarter. What matters is the market’s opinion of a sustainable long-term margin rate.

There was also an analyst who pointed out weak guidance for next quarter and asked whether there would be a reacceleration afterward, it was phrased like this:

…I wanted to ask Colette and Jensen with regard to sequential growth. So very strong sequential growth this quarter and you’re guiding to about 7%. Do your comments on Blackwell imply that we reaccelerate from there as you get more supply?…

Huang’s answer was very curt, in that the company guides one quarter at a time. Again, not reassuring there wouldn’t be further issues. This was followed up by CFO Kress:

…We are working right now on the quarter that we’re in and building what we need to ship in terms of Blackwell. We have every supplier on the planet working seamlessly with us to do that. And once we get to next quarter, we’ll help you understand in terms of that ramp that we’ll see to the next quarter and after that…

She specifically mentions suppliers, which indicates to me, it may have been challenging to get everything lined up for the next quarter. It’s not being said here, but it could be that there’s still some uncertainty (obviously depending on demand as well) what the supply situation would be afterward. Neither reassures me it will be smooth sailing. That’s not necessarily required for Nvidia to do well, but it is something that stood out to me from the call.

Huang also briefly addressed the issue of the new administration and whether this would impede Nvidia’s ability to sell into China. I’ve never heard Huang so firmly state that the company would support the administration:

…Whatever the new administration decides, we will of course support the administration. And that’s our — the highest mandate. And then after that, do the best we can.

…And so we have to simultaneously and we will comply with any regulation that comes along fully and support our customers to the best of our abilities and compete in the marketplace. We’ll do all of these three things simultaneously…

In my view, China isn’t a huge issue. NVIDIA still sells into China, but the business is a fraction of what it has been. In the past, it was a larger threat. I mostly view China as upside if the U.S. China relations were to defrost over time.

Huang closed the call by addressing the new age of AI:

…There are more AI-native start-ups than ever and the number of successful inference services is rising. And with the introduction of ChatGPT o1, OpenAI o1, a new scaling law called test time scaling has emerged. All of these consume a great deal of computing. AI is transforming every industry, company, and country. Enterprises are adopting agentic AI to revolutionize workflows. Over time, AI coworkers will assist employees in performing their jobs faster and better. Investments in industrial robotics are surging due to breakthroughs in physical AI…

I tend to be skeptical of new technology and its impact. With AI, I feel its impact is still underestimated by many people. I have no doubt NVIDIA has a lot of revenue runway ahead. What CEO Huang gets at with “time scaling” is the idea for models to filter our input (questions or whatever) and take more time when it is needed. In other words, be flexible about the amount of computing power that’s expended. This should translate into more computing power being drawn. Enterprise AI is big, of course, and has been relatively slow to take off because there are myriad data, privacy, security, legal, and other issues that have to be navigated. This is one of the reasons Palantir’s CEO Alex Karp believes that the money is really in the workflow and application layer.

With all these AI-native startups launching, we’ll see an incredible amount of specialized AI-models over the coming years. The interesting thing is that this specific technology isn’t rate-limited by our attention. Regarding media or food (just some examples), there’s only so much we can consume. AI models can interact with other AI models. General models can interface with us and in the background kickstart all kinds of specialized models depending on the context. This doesn’t require our attention at all. We’ll outsource more and more tasks to general models as long as the cost isn’t prohibitive. It seems obvious to me that hardware by Nvidia and Advanced Micro Devices (AMD) are key to continuing to lower the cost of inference and thus the cost of using AI models in the background wherever possible. Another potential medium-term driver of AI CapEx could be the scale-up of autonomous driving. I recently pointed out Alphabet is scaling up Waymo.

Nvidia has been trading rich for years. It is trading at 50+ forward earnings, versus a sector average of 27x. The sector average is probably skewed high because some chip companies are barely profitable. If we look at EV/Forward sales, it is trading at 27x. That’s compared to the sector average of 3x EV/forward sales. Consensus estimates have generally been revised up over the past few years. That’s often the case. Still, analysts are factoring in a slowdown in EPS growth in a couple of years:

It’s possible if the company loses market share. But if NVIDIA can continue to drive down power consumption/compute, it is difficult to envision its sales slowing down so early.

Although largely a good month for the market, December has been the worst month for Nvidia over the past 10 years. On a monthly basis, overall, Nvidia has on average registered a positive return 64.53% of the time. That’s excellent, of course, and could be expected of what must be one of the best-performing stocks over that timeframe.

However, Nvidia stock registered a positive return in December only 40% of the time. On an overall monthly basis, Nvidia on average returned an astounding 4.94%. That’s unbelievably good, but it turns to a negative 1.8% if we just look at December. Given December is generally a solid month for stocks, this difference is quite astounding to me. Interestingly, the difference isn’t because NVDA registered some abnormal mammoth momentum crash in December. It simply registers negative often and didn’t put up any big positive results. Its best December goes back to 2016, when it put up 15.8%. Could be different this time, of course.

Based on the seasonality data, for the next month, I don’t really like Nvidia Corporation stock right now. In my previous Nvidia article, I rated it a hold. I’m going to stick with that because the medium and long-term story is still intact. Blackwell’s problems are most likely just muddying near-term revenue. What is significant for the stock is its continued long-term market share dominance and growth of the demand for AI. As much attention as it gets, I suspect long-term AI demand is widely underestimated.

---

Analyst’s Disclosure: I/we have no stock, option or similar derivative position in any of the companies mentioned, and no plans to initiate any such positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Seeking Alpha’s Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.

---