Nvidia may be one of the most popular artificial intelligence (AI) stocks around, but have you ever thought about which companies are buying its graphics processing units (GPUs)? Big tech companies make up a lot of the consumption, but they aren't always using them for their own computing power.
Instead, many of these GPUs are going into massive servers that are available for rent. Customers can rent out whatever computing power they need to run AI models, store data, or process information as needed. This is a smart move, as most of these customers don't need massive computing power available at all times.
So, instead of Nvidia, it would be smart to invest elsewhere in the AI value chain -- in some of the largest cloud computing providers.
The cloud computing business model is tried and true
The biggest cloud computing providers are Amazon (AMZN 0.55%), Microsoft (MSFT 0.07%), and Alphabet (GOOG -0.92%) (GOOGL -1.02%). With their Amazon Web Services (AWS), Azure, and Google Cloud offerings, they control about two-thirds of the cloud computing market.
Their business models are also quite attractive compared to Nvidia. When Nvidia sells its GPU, that's it. It's a one-time sale, and the only way Nvidia keeps making more money is to replace the GPU after its service life is completed. That can be a few years, so Nvidia's product demand may be a bit lumpy.
Cloud computing providers have a one-time expense to buy the GPUs and build the servers, then have a relatively straightforward operating cost of electricity and general maintenance. Then, the rental revenue kicks in and more than pays off the upfront cost.
To simplify things, you could think of Nvidia as someone who provides all the materials to build an apartment complex and the cloud computing providers as businesses that assemble the pieces and rent out the complex as landlords. This analogy isn't meant to be pushed too far, but it's a good way to break down how it works. Over time, the landlord makes far more money than the company selling supplies, but both are required to make the business work.
This is what makes cloud computing providers such fantastic investments. The problem is that they can't be purchased on their own.
Each company's cloud computing segment is excelling
Cloud computing is only a fraction of each company's business.
AWS only accounts for 18% of Amazon's sales, with about $25 billion in revenue in the first quarter. However, AWS' margins are much higher than those of the commerce side. In Q1, Amazon posted $15.3 billion in operating profit, but 62% of that came from its AWS business. AWS is the largest cloud provider of the trio, with an estimated 31% market share, according to Synergy Research Group.
However, Microsoft is swiftly gaining ground. In the period ended March 31 (Microsoft's third quarter of fiscal 2024), Azure revenue grew 31% year over year. Unfortunately, Microsoft doesn't break out how much revenue Azure generates, but it does report information on its Intelligent Cloud division, of which Azure is a part.
We can piece together an estimated revenue figure. In Q3, Intelligent Cloud had revenue of $26.7 billion, and during the fourth quarter of fiscal 2023 (ending June 30, 2023), Azure made up more than 50% of Intelligent Cloud's revenue for the first time. This share has likely increased since then, so it's safe to say Azure's revenue was likely around 60% of the total, or about $16 billion.
In last place is Google Cloud, which saw revenue growth of 28% to $9.6 billion. Google Cloud is still working on optimizing profitability and only posted an operating margin of 9%, although this metric has steadily grown. While Google Cloud posted about a third of revenue as AWS, don't write it off. Google Cloud boasts a significant concentration in workloads centered on AI, with 60% of funded generative AI start-ups and 90% of generative AI unicorns (private companies valued over $1 billion) being Google Cloud customers.
All three companies are excellent investments. As AI becomes more integrated into the business, the cloud computing providers will do a lot of the heavy lifting. This makes them great investments and an excellent alternative to investing in Nvidia.