The AI Evolution

The recent surge in Artificial Intelligence (AI) isn't novel. The term "AI" first emerged in the 1940s for decoding wartime messages and entered popular consciousness when IBM's Deep Blue defeated chess legend Garry Kasparov in 1997. AI became more personal with the launch of Apple's Siri in 2011 and the widespread adoption of Amazon's Alexa in 2014. Fast forward to the 2020s, and we see the introduction of ChatGPT. While ChatGPT quickly gained traction, amassing 100 million users in just two months, its usage has seen a significant decline. Is AI just another tech trend?

Many argue that AI's current iteration is more than just a passing phase, pointing to its repeated cycles and the extensive business adoption observed recently. With the consensus that AI is a permanent fixture, investors are eager to discover the pivotal players in this AI gold rush. The answer? Nvidia and their H100 chips.

Since its founding in 1993, Nvidia has led the industry in GPUs (Graphics Processing Units), transitioning from supporting the computer gaming sector to enabling high-throughput computing, crypto mining, and now, Artificial Intelligence. Nvidia's historical performance and market dominance make it a preferred choice, leading to record earnings and stock prices after their Q2 2023 earnings report. They've reported a 101% increase in revenue this year, with operating income soaring by 1263%. Currently, Nvidia boasts a market cap of $1.22 trillion, positioning it alongside tech behemoths like Google, Amazon, and Microsoft.

Nvidia's exponential growth is remarkable, reminiscent of Sun Microsystems' trajectory. Once a leading computer component manufacturer known for its superior chips, Sun Microsystems couldn't sustain its edge over time. As the fervor around Web 2.0 waned and rivals introduced competitive technologies, Sun Microsystems became just another player in the industry.

Yet, a recent development raised eyebrows. During their earnings call, Nvidia highlighted a $2.3 billion earnings boost from data center sales. Intriguingly, this aligns with the Line of Credit (LOC) Nvidia extended to the emerging AI Cloud leader, CoreWeave. Backed primarily by BlackRock, CoreWeave utilized this LOC to purchase $2.3 billion worth of Nvidia H100 chips. This move not only fortified Nvidia's revenue figures but also ensured that should CoreWeave falter, the chips would return to Nvidia. What makes this more fascinating is BlackRock's significant stake in Nvidia - owning 182 million shares. Following the earnings report, Nvidia's stock surged by 11%, augmenting BlackRock's holdings by an estimated $9.4 billion. Considering the initial $2.3 billion LOC, the potential ROI (Return On Investment) is staggering, even more so when acknowledging that BlackRock wasn't the sole contributor.

Such maneuvers, while financially astute, echo the speculative strategies employed during the Dot-Com Bubble, such as those seen with Sun Microsystems.

While no one anticipates Nvidia's decline in the immediate future, assuming they'll continue to monopolize the market is a stretch. As technology evolves, many corporations prefer custom solutions to control their infrastructure. Google, for instance, designed bespoke servers and silicon systems to buttress its vast ecosystem, including YouTube. Amazon followed suit to optimize AWS (Amazon Web Services). Tesla, drawing from its autonomous driving data, crafted its own chips. The question isn't if GPUs and AI infrastructure will become ubiquitous and commoditized - it's when.