No company embodies the artificial intelligence (AI) age more than Nvidia (NVDA 6.18%), whose GPUs (graphics processing units) are being used in data centers to power AI applications.
Nvidia's growth has been extraordinary. The company grew its revenue an incredible 265% in Q4 to $22.1 billion, with data center revenue soaring 409% to $18.4 billion. And with Nvidia's growth, investors have benefited, with the stock more than tripling over the past year.
Image source: Getty Images.
The AI revolution has led to comparisons to the internet boom and bust. Not surprisingly, the company Nvidia is most compared to is Cisco (CSCO -0.50%), which was the backbone of the internet.
Cisco became the world's most valuable company in 2000, but its stock crashed soon afterward and Cisco investors never fully recovered. The stock still trades well below its all-time high today.
So, can Nvidia and its investors avoid the same fate as Cisco investors?
There are similarities but also key differences
In many ways, the rise of Nvidia is reminiscent of Cisco back in the late 1990s and early 2000s. During that period, Cisco was the leader in providing the infrastructure equipment used to power the internet boom through its network routers, much like Nvidia is powering the AI revolution through its GPUs today.
And like Nvidia, Cisco in 2000 was primarily a hardware business. This is important because software revenue is more recurring and higher margin. As a result, investors typically assign higher multiples to software businesses. Neither Nvidia nor Cisco are primarily software businesses, and their multiples should reflect this.
While Nvidia today and Cisco in 2000 share these similarities, there are key differences. One is that Nvidia's Cuda software platform has been a key differentiator for the company since it was introduced in 2006, even though Nvidia gives it away for free.
This software platform allows Nvidia GPUs to be programmed directly, saving customers time and money. By the time alternatives evolved, Nvidia GPUs had already become the industry standard. This created a wide moat for the company. Investors like wide moat businesses as these companies face less competition.
Cisco in 2000 was just starting to get more meaningfully into software with the introduction of its CiscoWorks2000 network management software when its stock peaked. Today, software is about a third of its revenue.
Nvidia's and Cisco's customer bases are also quite different. Cisco was heavily selling into the cable and telecom industries, where companies typically carry meaningful leverage and have unpredictable capital expenditure (capex) cycles. Capex is the purchase of assets meant to be used over a long period. Nvidia is selling into the data center, where cash-flush companies like Amazon, Microsoft, and Alphabet preside. This gives Nvidia a more stable customer base, which is another positive for investors.
Nvidia also trades at a much lower valuation than Cisco did at its peak in March 2000. At its highs, Cisco had a price-to-earnings ratio (P/E) of nearly 200. .
Nvidia, meanwhile, trades at a P/E of around 60, and has been growing its revenue more quickly. In 2023, Nvidia's revenue surged 126%, while Cisco's revenue rose 55% in its fiscal year 2000. Investors generally assign larger multiples for high revenue growth, and there are few companies growing as quickly as Nvidia.
Potential pitfalls
The biggest risk Nvidia faces is one that eventually hurt Cisco: demand suddenly drying up after the initial order surge. In 2001, Cisco took a large inventory write-off because in the race to meet insatiable demand, it acquired too much inventory and then demand suddenly dropped off. This was essentially Cisco saying the components it purchased to build its networking equipment were now basically worthless.
Currently, there seems no end in sight for GPU demand as companies build out their AI infrastructure. However, Cisco in 2000 believed there was no end in sight to the demand for networking routers. That outlook did not end well for Cisco investors.
Nvidia also has a bit more customer concentration than Cisco did in 2000, with one indirect customer representing 19% of sales in 2023. Customer concentration adds more risk, because if a large a customer pulls back on spending, it can lead to a negative earnings surprise, which tends to put pressure on stocks.
Still early innings
Nvidia is currently one of the best growth stories in the market. While there are similarities between it and Cisco 2000, they are also key differences. Nvidia is also much cheaper than where Cisco was when it peaked and it's been growing faster. Investors need to be aware of the risks, but AI still appears to be in the early innings. After all, AI wasn't a huge topic of conversation a few years ago, and now nearly every company is talking about it. Investors can continue to ride Nvidia's great growth, but just be aware it might not last forever.
