Semiconductor designer Nvidia (NVDA 6.18%) emerged as an artificial intelligence (AI) titan in 2023, capturing investors' imaginations with its groundbreaking advancements and a stock price that has soared to remarkable heights. But with great prominence comes great scrutiny.
As the discourse around Nvidia's valuation intensifies, three of The Motley Fool's finest tech writers have gathered to dissect the narrative surrounding Nvidia. Their discussion explores the landscape of Nvidia's valuation, its pioneering role in the AI domain, and the formidable competition on the horizon.
Together, this diverse set of Nvidia insights should help you set your own course in these uncharted waters.
Generative AI isn't just hype, and Nvidia's lead is the real deal
Nicholas Rossolillo (Bullish on Nvidia): Generative AI proves that something can both be hype and the real deal all at once. Just a year out from ChatGPT becoming a viral hit, Nvidia's pioneering work in GPUs, networking chips, and software frameworks that power large-language models -- the basis for generative AI systems -- is paying off in spades.
The company's data center segment was a high-growth unit before, but it's now doing things at scale never seen before. Over the summer of calendar year 2022, Nvidia's quarterly data center revenue was just shy of $4 billion. This summer, it soared past $10 billion in a single quarter, and CEO Jensen Huang and the top team have indicated that number will head higher.
Of course, it's still early days for this new iteration of AI, but early estimates from researchers and from the semiconductor industry itself (like chipmaking giant Taiwan Semiconductor Manufacturing (TSM 1.26%), a key Nvidia production partner) agree. The generative AI market could experience growth of over 30% per year on average through the end of this decade, in which time it could sop up many tens of billions of dollars in annual spending.
But what of Nvidia's valuation? The company is lapping a period last year in which profitability was close to nil. On a forward-looking basis, shares could be reasonably valued, especially if Nvidia continues to clobber all growth and cash-generation expectations as it has been.
Data by YCharts.
That being said, Nvidia stock has been highly volatile in the past. Its business is cyclical, with new applications for its chips and software leading to some wild boom-bust cycles. As bullish as some of us might be, this fact needs to be kept in mind, and enthusiasm should be checked. I fully expect Nvidia to go through more turbulence at some point when generative AI encounters growing pains. In the meantime, I like select software stocks such as ServiceNow (NYSE: NOW) that could get a growth boost from working with Nvidia.
Nevertheless, new AI services are absolutely the real deal, and Nvidia's leadership won't be easy for semiconductor competitors to catch up to. Nvidia is my largest stock holding right now, and while I'm not buying more, I'm certainly not about to sell either.
Nvidia has huge growth potential, but also huge competition coming for it
Billy Duberstein (Hold Nvidia): I find myself vacillating between thinking Nvidia is a terrific buy and being fearful of buying at these levels.
On the one hand, Nvidia has a massive lead on would-be competitors, by having not only the best general-purpose GPU for parallel processing, but also a full AI tech stack enabled by its CUDA programming language. Nvidia has a significant lead here, having begun CUDA in 2006.
Obviously, Nvidia's massive multiyear investment is paying off this year. Data center revenue surged by a shocking 171% last quarter, and management has guided for much more growth this year. And Nvidia's operating margins also expanded to a sky-high 50%.
But if you think that's impressive, you ain't seen nothing yet. Most key industry participants, from Advanced Micro Devices (AMD 2.37%) CEO Lisa Su to Nvidia's foundry Taiwan Semiconductor Manufacturing, predict that the growth of AI chips will be around 50% annually over the next five years. To put that in perspective, compounding growth at 50% for five years is more than 7.5x growth. That's a massive number.
Of course, at 111 times earnings, Nvidia's stock price is certainly giving the company credit for some -- but probably not all -- of that growth. That means Nvidia has to keep its lead ahead of not just Advanced Micro Devices and Intel (INTC -9.20%), but also every large cloud computing company that's also developing an in-house accelerator. Moreover, these companies are likely to settle on a common open source programming code that will work across a variety of AI chips.
The common cause for open-source AI software and financial firepower among these challengers may pose a threat to CUDA. However, it appears Nvidia is already on its way to one-upping that, as it recently released a new innovation cadence in its October investor presentation. Nvidia is now moving to a one-year cadence in introducing new architecture, whereas before it was releasing new chips on a two-year cadence. with the A100 released in 2020 and the H100 in 2022.
The new hyperspeed pace could push Nvidia even farther ahead, but it will also take investment in research and development (R&D). In addition, hardware superiority doesn't provide as large a moat as software network effects.
Because Nvidia has such amazing potential, but also a high multiple and such well-funded competition on the way, that leaves me somewhere in the middle. I'd love to own the stock, but at a lower price. However, I also wouldn't caution others against buying at these levels, as the AI boom appears to have legs.
It's tempting to cash in on Nvidia's sky-high valuation
Anders Bylund (Nvidia looks overpriced): I'm only human, and my investing strategy is far from perfect. The biggest hole in my skill set is a reasonable sense of when it's time to sell. Whatever stocks, exchange-traded funds (ETFs), and cryptocurrencies I might buy, I invariably aim to hold for many years. Master investors Benjamin Graham and Warren Buffett say that the best holding period is "forever," and I take that idea seriously. Too seriously, from time to time. There have been countless instances where I missed reaping substantial benefits by holding on to a winning investment long after its momentum had waned.
Against that background, you should know that I own a few Nvidia shares and am sorely tempted to cash in some of my winnings. I have no plan to liquidate my entire Nvidia position, but selling a couple of shares would at least lock in some of my gains so I don't have to worry about losing a game-changing chunk of it in a drastic price correction.
My crystal ball is in for a tuneup so I can't tell exactly when Nvidia's stock chart could turn sharply downward, how far the plunge will go, or whether it's happening at all. I've been wrong before and most Nvidia investors would love to see me miss the mark in this case, too. Maybe the best strategy is to sit on my hands and leave the sell button alone -- but I see too many signs of an overpriced stock that seems overdue for some sort of correction.
I do get the investor excitement. Nvidia's chips were instrumental in setting up OpenAI's ChatGPT for success, and that big win is bringing a plethora of follow-on sales from every corner of the tech sector. The company has a valuable first-mover advantage here, and that's why Nvidia's stock has skyrocketed over the last year.
But this early price spike could be a bullish overreaction, destined to fade out and turn into at least a temporary price drop just as quickly.
Nvidia is far from the only name in the AI chip game. Archrival Advanced Micro Devices also offers processors specifically designed for the training phase of developing machine learning systems. Intel has a neural network accelerator called Gaudi, and the Goya processor for the learning stage. AI platform operators Alphabet (NASDAQ: GOOG) (NASDAQ: GOOGL) and Amazon (NASDAQ: AMZN) have also designed their own AI chips, even though they mainly focus on software. So Nvidia has an early lead, but who knows which solution (or solutions) may dominate the next couple of years -- or the long haul?
Optimistic investors appear to have priced the AI explosion into Nvidia's current stock price -- they incorporated the entire opportunity and then some. Nvidia's stock is trading at a nosebleed-inducing 32 times sales and 102 times trailing earnings. Nvidia must execute its AI plan to perfection in order to support these lofty stock prices, and I'm not convinced that there's a reasonable business plan strong enough to get the job done.
So I've considered selling about one-third of my Nvidia holdings while the getting is good. With a 160% gain since I picked up those shares in the summer of 2022, that move would return nearly the entire original investment to my cash reserves. That cash would be put to work in lower-risk investment ideas while the other two-thirds portion can stay committed to the AI boom.
This is not like the time I cashed in 90% of my Tesla (NASDAQ: TSLA) investment, because I have not lost faith in Nvidia's excellent products and brilliant leadership. But I'm definitely not buying new Nvidia stock in this environment, and those paper profits would definitely be more helpful if converted into solid cash gains instead.
And as I said, this kind of thinking is rare for me; I wouldn't be thinking about stock sales at all if Nvidia's valuation was a little more realistic. But here we are, with the stock soaring in the clouds -- and right now I'm quite interested in skimming some profits off the top.
