Pay Attention: Musk xAI-Nvidia Circular Deal, Oracle Report Shows Lower Margins on Renting Nvidia GPUs
News Summary
A new report indicates that Oracle Corp generated only a 14% gross margin on $900 million from renting Nvidia GPU-based servers over the last three months, significantly lower than credible analyst expectations of 25% and "momo guru" projections exceeding 50%. This raises concerns that profit forecasts for AI infrastructure, especially for lesser-known players, may be overblown. The article also highlights an accelerating trend of "circular financing" within the AI sector. For instance, Nvidia is investing $2 billion in Elon Musk's xAI, which will then buy Nvidia chips. Previously, Nvidia invested in OpenAI, which subsequently purchases chips from Nvidia and rents AI servers from Oracle, while Oracle itself is a Nvidia chip customer. This pattern is reminiscent of the circular financing that partly contributed to the 2000 tech stock crash, which saw the Nasdaq lose 78%. Despite these "red flags," the author reiterates a firm conviction in the long-term AI opportunity, asserting that fortunes are to be made in AI through 2030, albeit with treacherous periods. The piece also briefly touches on money flows in Magnificent Seven stocks and advises investors to consider protection bands (cash or Treasury bills) to manage risk while participating in upside potential.
Background
The global investment community currently holds immense enthusiasm for Artificial Intelligence (AI), particularly in high-performance computing and AI infrastructure. The business of renting computational power, primarily centered around Nvidia GPUs, is seen as a nascent market with significant growth potential. Many companies are aggressively investing in AI servers and data centers, anticipating substantial profits from providing AI compute services. However, market expectations for AI infrastructure profitability have been exceptionally high, with some "momo gurus" projecting gross margins exceeding 50%. Historically, the tech industry has witnessed market bubbles fueled by "circular financing," such as during the dot-com bust of 2000, where companies inflated revenues and valuations through reciprocal investments and service purchases, ultimately leading to market collapse.
In-Depth AI Insights
What do Oracle's lower-than-expected margins imply for AI infrastructure investment? - Oracle, a well-capitalized player with Larry Ellison's significant connections, achieving only a 14% gross margin on GPU rentals suggests that the actual profitability of AI infrastructure businesses might be substantially lower than market expectations. - This indicates that the costs associated with acquiring AI hardware, operational overhead, and market competition are likely more intense than anticipated, severely compressing profit margins. - For smaller, less connected AI infrastructure rental companies being hyped by "momo gurus," projections of 50% or even 25% margins are highly likely to be overstated, posing significant investment risk. What systemic risks does the accelerating circular financing model in AI present? - Circular financing artificially inflates reported revenues and capital flows within the sector, leading to a distorted perception of genuine demand and profitability. The reciprocal investments and purchases between Nvidia, xAI, OpenAI, and Oracle mean the same dollars are counted multiple times, obscuring true financial health. - This pattern bears a striking resemblance to the conditions preceding the 2000 tech bubble burst, where companies like Lucent and Nortel engaged in similar financing schemes, ultimately leading to market collapse. It creates a self-reinforcing valuation bubble that is prone to rapid deflation once external capital or genuine user demand fails to sustain it. - Such unsustainable capital recycling will eventually expose companies lacking independent profitability and a real customer base to liquidation risks, potentially dragging down valuations across the entire AI sector. How should investors balance a belief in AI's long-term potential with current market "red flags"? - Investors must differentiate between the long-term transformative potential of AI technology and short-term market speculative behavior. While AI remains a powerful secular trend through 2030, the investment journey will be volatile. - Given the warning signs of low margins and circular financing, investors should prioritize fundamental analysis, focusing on genuine customer acquisition, and sustainable business models rather than merely chasing concepts or high valuations. - Adopting a protection band (e.g., cash or short-term Treasury bills) is prudent, allowing investors to preserve capital during potential market corrections while retaining flexibility to invest when more reasonable valuations or new opportunities emerge in the AI space.