Nvidia Q3 Preview: This Is The Big Quarter

Summary:

• Nvidia Corporation’s stock has surged 31.32% driven by optimism around the Blackwell GPU, but I believe this enthusiasm is overly optimistic and exposes valuation risks.
• The Blackwell GPU faces overheating issues in server configurations, complicating its adoption and potentially increasing costs for hyperscalers, which could impact Nvidia’s sales.
• Despite upward revisions for Q3 earnings, Nvidia’s high valuation and reliance on Blackwell’s success make its current stock price unsustainable, warranting a strong sell recommendation.
• Hyperscalers’ existing investments in older Nvidia chips and potential for achieving AGI with current hardware reduce the urgency for Blackwell adoption, posing further risks to Nvidia’s growth.

Investment Thesis

Nvidia Corporation (NASDAQ:NVDA) shares are up 31.32% since my last update on the AI chip giant in September. Nvidia’s powerful gains (while I think are unwarranted) have been driven by continued investor optimism surrounding the expected performance of their Blackwell GPUs.

The Blackwell GPU has sparked strong market excitement that these new chips could be Nvidia’s new super product that accelerates the company to the next leg of growth, as the California company continues to dominate the AI GPU industry (for now).

While their dominance is undeniable, I’m still bearish. The investor expectations surrounding Blackwell’s are still overly optimistic, which means shares are exposed from a valuation standpoint.

I think there’s little doubt that Nvidia’s stronghold in GPUs will continue for the short term, but with their stock priced at a premium, investors are likely overly enthusiastic about the potential for growth.

I think it is difficult to see how they can continue to maintain such high growth rates, especially with news out this weekend that their new chip line overheats in some server configurations. This is a big deal because it means hyperscalers have to incur extra costs to incorporate these chips into their data centers.

All eyes will be on Blackwell’s performance heading into Q3 earnings post-market on November 20th. Given the hyper bullish sentiment, I continue to believe that Nvidia shares are overextended and remain a strong sell in the short term.

Why I’m Doing Follow-Up Coverage

In my previous research piece on Nvidia, I really focused on my personal concerns about the sustainability of the AI giant’s growth, especially given the delays (at the time) surrounding their Blackwell chip. I argued it would be necessary for sustaining sequential revenue growth in North America. The lack of Blackwell’s timely rollout likely contributed to a sequential revenue decline in the US last quarter.

Nvidia’s current market valuation feels inflated and does not appear to reflect the risks associated with its delayed product rollout or issue with the product itself. I continue to believe excitement surrounding Blackwell may be overblown.

While the company is said to be ramping up production for Blackwell with partner TSMC (TSM) and announced this past week that Japan’s SoftBank will be receiving the first chip, I think there are product issues that remain. Q3 earnings is a big deal for the trajectory of shares.

Even CEO Jensen Huang admitted to the initial delay (and said it was his company’s fault) during an event in Copenhagen:

We had a design flaw in Blackwell….It was functional, but the design flaw caused the yield to be low. It was 100% Nvidia’s fault.

This is why I’m conducting follow-up coverage. I continue to believe the market’s bullish stance on Nvidia overlooks these structural issues.

With Blackwell Overheating, This Is A Big Quarter

Before we dive into why this is a big quarter, I want to touch on what the implications are of Blackwell overheating.

This weekend, The Information reported (and Seeking Alpha added research to) news that Blackwell chips were overheating when placed in larger clusters in datacenters.

According to the report, the Blackwell graphics processing units overheat when they are connected together with server racks that are designed to hold up to 72 chips, sources familiar with the issue said.

For hyperscalers, this is a big deal. GPUs need to be placed closely together in order for their datacenters to run enough compute per square foot of floor space for the datacenter investment to make sense. If datacenters need more space to accommodate these new chips, this will be a big setback for many of Nvidia’s customers.

With this, it was already an important quarter for Nvidia, considering the company reported a sequential revenue decline in the US last quarter, and this performance signals a potential shift in demand from hyperscalers in the US (the largest market for Nvidia’s products).

Adding another layer of complexity to this quarter is whether hyperscalers even require the new chips. OpenAI CEO Sam Altman has claimed that achieving Artificial General Intelligence (AGI) with existing hardware is possible​. If AGI can be achieved with current GPUs, the need for constant, exponential hardware upgrades may be less pressing. This is the heart of Nvidia’s business model.

During a recent Goldman Sachs conference, CEO Huang explained:

I would say several things that are very different about us. The first thing is to remember that AI is not about a chip. AI is about an infrastructure. Today’s computing is not build a chip and people come by your chips, put it into a computer, that’s really kind of 1990s. The way that computers are built today, if you look at our new Blackwell system, we designed seven different types of chips to create the system. Blackwell is one of them.

Blackwell has a lot of promise. The question is if customers will want it as much as the market thinks, given the promise of existing hardware and the specific needs these new chips have.

Q3 Preview

As the tech giant prepares to report Q3 earnings on November 20th after the bell, analysts have been mostly revising their estimates upward over the last three months.

For the fiscal year ending in January, there have been 46 upward revisions and just 4 downward revisions for revenue, and 40 upward revisions compared to 6 downward revisions for earnings per share.

EPS for the quarter is expected to come in at $0.74/share, with revenue of $33.07 billion. This would represent EPS growth of 85.14% YoY, and revenue growth of 82.50% YoY, respectively.

To me, while these expectations are likely attainable (because they do not rely on Blackwell demand), Q3 is expected to be the last quarter without Blackwell sales. This means investors (and analysts) will need to see on the call that Blackwell revenue will be booked in FY Q4 (the current quarter) to support full-year expectations.

On the earnings call, I’ll be looking for data on how Nvidia is solving their Blackwell technical challenges. I think it’s clear this is the single most pressing problem facing the company right now.

Valuation

I believe Blackwell demand is not as solid as it appears, given the strength of current hardware and the caveats that come with buying the new GPUs. With this, Nvidia’s valuation remains a point of concern to me. The company trades at an expensive forward P/E ratio of 49.75 for the fiscal year ending in January 2025​. This compares to the sector median forward P/E of 24.25 times forward earnings.

On a price to sales standpoint, the company is expensive as well. Shares trade at 27.60 times forward sales. This is an 820.16% premium over the sector median forward price to sales of 3.00.

Personally, I don’t see how either of these premiums are sustainable.

The introduction of the Blackwell chip, while crucial for certain use cases, may not experience the same widespread adoption as previous chips like the H100 and H200, which had more appeal across the gaming, data center, and AI industries. The H100 and H200 had blue ocean level opportunities. I don’t believe Blackwell does, especially given Sam Altman’s comment.

Given this, (and I know this is controversial), I firmly believe Nvidia should be trading at the sector median forward P/E ratio. Currently, the forward P/E is 24.25 I mentioned before. If we saw Nvidia’s shares converge on this 24.25 times forward earnings, shares would have a potential 51.25% downside from their current position.

Bull Thesis

To be clear, I am not trying to diminish the value the new Blackwell chip can bring to some hyperscaler customers. The new GPU still has a pivotal role to play in advancing AI model inferencing, an area where performance is everything. According to CEO Jensen Huang, it could be up to 50 times faster on some inferencing tasks.

For some hyperscalers, Blackwell offers a compelling solution for companies that absolutely need top of the line inferencing capability. They clearly are still an attractive proposition for industries heavily dependent on AI.

Again, as I mentioned before, Blackwell’s adoption will not be without challenges. Over the last 24 months, companies, including major software firms like Microsoft (MSFT), have already made massive capital expenditures on AI infrastructure, including GPUs. These large investments were designed to support previous generations of Nvidia chips, such as the H100 and H200.

Asking these big hyperscalers to change their server infrastructure due to an unforeseen overheating issue with the Blackwell chips is a big setback for Nvidia. With such heavy investments already in place (and because Altman identified that older chips like the H100/H200 can power AGI) I think the prospect of abandoning older tech in favor of the Blackwell chips seems far less likely than before.

Keep in mind that for these software tech giants, these massive capex projects have caused a big shift in their business models from high-margin software businesses to capital-intensive infrastructure investments. Already, this creates friction for these tech giants, and they are incentivized to make current investments last as long as possible. Fellow Seeking Alpha analyst Bluesea Research wrote some solid analysis over the summer showing how capex announcements have spooked Wall Street.

Even though Blackwell’s AI capabilities are undeniable, companies will likely take a measured approach to integrate this new chip. I don’t think the Street is expecting a measured approach.

Takeaway

Nvidia’s stock has been on an incredible upward trend, driven by investor enthusiasm surrounding the rollout of the Blackwell chip, since my last coverage. However, while the technology is promising, I think investors need to be more cautious than they currently are. The market has already priced in incredible growth, and I think there is a real risk that the overheating caused by the Blackwell chips could mean sales from traditional hyperscalers will disappoint.

If Blackwell doesn’t meet the market’s high expectations, Nvidia’s valuation will likely face downward pressure. Hyperscalers have already made substantial capex investments, and companies are incentivized to make the most of these investments first. While I have been consistently wrong with my analysis on Nvidia stock over the past year, I do believe we are close to demand growth slowing for GPUs from the tech giant. With this, I think shares continue to be a strong sell.

Editor’s Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.

---

Analyst’s Disclosure: I/we have a beneficial long position in the shares of TSM either through stock ownership, options, or other derivatives. 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.

Noah Cox (main account author) is the managing partner of Noah’s Arc Capital Management. His views in this article are not necessarily reflective of the firms. Nothing contained in this note is intended as investment advice. It is solely for informational purposes. Invest at your own risk.

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.

---