NVIDIA: No Bubble Yet
NVIDIA is selling the tools for the AI revolution, and every industry will gradually adopt AI, just as the software industry has. This would allow NVIDIA to not only maintain its current revenue growth rate, but even increase it.
This is the author's opinion only, not financial advice, and is intended for entertainment purposes only. The author holds a beneficial long position in NVIDIA Corp.
Taiwan as the world's production center for high-end chips and especially NVIDIA's GPUs reminds me of the desert planet Arrakis in Frank Herbert's Dune. Just as the spice necessary for interstellar travel can only be produced on Arrakis, the chips necessary for Large Language Models (LLMs) can only be produced in Taiwan. Since NVIDIA's H100 chips enable artificial intelligence in all its applications, they have become a geopolitical resource, with Saudi Arabia or the United Arab Emirates, for example, buying them at scale for strategic reasons. Although competitor AMD recently made progress with the release of its MI300 chips, NVIDIA is still ahead of the race. NVIDIA's H100 remains the gold standard - the spice - for AI infrastructure and is still superior to AMD's MI300. As a result, NVIDIA maintains an economic moat that it can even widen through massive investments in research and development.
Given the current hype around AI, it is understandable that investors are reminded of the dot-com bubble, but it may be far too early for that. According to George Soros, stock market bubbles are much more complex than simply inflated price-earnings ratios. After all, it was not a misjudgement of the future impact of the Internet in the late 1990s that led to a bubble, but rather a misjudgement of the time frame in which the Internet could actually increase the productivity and efficiency of our economies. However, the impact of generative AI on businesses and our economy may become measurable much more quickly. The most obvious area of early adoption of generative AI to date has been in computer programming and software development, where it is already happening at breathtaking speed. As early as July 2023, experts at McKinsey estimated that AI could double the speed of production in the software industry. Thousands of companies are already using Github Copilot, and more than a million users are paying for the service. NVIDIA CEO Jensen Huang recently predicted "that the programming language is human, everybody in the world is now a programmer. This is the miracle of artificial intelligence" - coding is dead. AI will enable a whole new way of interacting with computers. But it's not just about assisting with coding: On March 5, for example, cybersecurity company Crowdstrike released its fiscal year 2024 results, reporting better than expected revenue and guidance. A key product is Charlotte AI, an artificial intelligence that democratizes cybersecurity management by enabling natural language interaction.
More exciting for NVIDIA's future, however, are the areas where generative AI has not yet been adopted at scale. For example, while there seems to be a consensus on the potential of AI, there has been little to no day-to-day application of AI in healthcare and medicine. We are still a long way from AI actually performing diagnostics, making diagnoses, interpreting medical imaging or histopathology slides, or even performing interventions on patients in the everyday world. AI could also soon revolutionize research in the chemical industry, simplify processes in old heavy industries, create all kinds of content for movies and video games or even turn government bureaucracies upside down. For example, the city of Heidelberg in southern Germany is already experimenting with an AI chatbot (Lumi by Aleph Alpha) to help people deal with government agencies and answer questions about the complex German bureaucracy. In my opinion, it is foreseeable that sooner or later AI will find its way into every niche of our civilization.
NVIDIA is selling the basic tools for this revolution, and every industry will gradually adopt generative AI, just as the software industry has. As if that were not enough of a tailwind, the next bitcoin halving is approaching (likely in April 2024). Mining bitcoins will then become more computationally intensive, which should further increase demand for NVIDIA's energy-efficient GPU chips. All of this could allow NVIDIA to not only maintain its current revenue growth rate (recently reported revenue growth was 126% year-over-year), but even increase that rate. Of course, that's hard to imagine with NVIDIA's current market cap of more than $2 trillion, or as physicist Albert A. Bartlett once wrote: "The greatest shortcoming of the human race is our inability to understand the exponential function." Remember what they say in the world of Dune? - "The spice must flow."