Podcast Summary
Nvidia earnings: Nvidia's earnings report fell short of expectations due to production issues with their new GPU, causing market concerns and a sell-off in Nvidia and related stocks
Nvidia's latest earnings report, which fell short of some optimistic estimates, raised concerns about the waning growth of the company. The issue wasn't with the design of their new GPU, Blackwell, but rather the production mechanism. The market wanted more details and reassurance from Nvidia's management, but they didn't provide precise information about sales figures and timelines. This lack of clarity led to a significant after-hours sell-off not just in Nvidia but also in its chip-making peers and server equipment providers like AMD and ARM. The market is seeking clarity and certainty in an uncertain economic climate.
Nvidia's earnings report: Despite missing some optimistic expectations, Nvidia's earnings report still beat estimates and came after significant gains. Valuation concerns and new product margins may have contributed to the after-hours decline.
Nvidia's after-hours decline following its earnings report should be viewed in the context of its strong year-to-date performance, which saw the stock gain over 150%. While the report did not meet the most optimistic expectations, it still beat estimates and came after a significant gain. The broader tech sector, particularly chipmakers and chip design companies, are experiencing widespread weakness following the report. However, the impact on the allocation of funds towards AI as a fundamental secular driver is unlikely to change significantly in the near term. The decline may be due to growing concerns about valuation and the fact that some stocks, including Nvidia, had been moving up significantly. Other factors, such as a short report on Super Micro, one of Nvidia's biggest customers, and its delayed 10K filing, may also be contributing to the weakness in Nvidia's stock. Gil Luria of A. Davidson, a relative bear on the stock, noted that while the quarter was strong, investors may be getting cautious about the margin situation, particularly with the introduction of new products like Blackwell, which put pressure on gross margins and caused less of the revenue upside to flow to the bottom line.
Nvidia's growth plan: Nvidia aims for $200B revenue by 2026, relying on data center biz and Blackwell, but production challenges raise doubts, strong customer relationships and complete systems offer advantage, hyperscalers remain major revenue source, but market may broaden
Nvidia's ambitious plans for reaching $200 billion in revenue by 2026 rely heavily on continued growth in their data center business and the successful ramp-up of their latest product, Blackwell. While the market's reaction to Nvidia's latest financial results was positive, some challenges with the production of Blackwell have raised questions about the feasibility of this aggressive growth plan. However, the company's strong relationships with key customers like Microsoft, Meta, Amazon, Google, and Tesla, as well as their ability to offer complete systems beyond just GPUs, give them a significant competitive advantage. The hyperscalers remain a substantial portion of Nvidia's revenue, but their consistent over-investment in AI capacity suggests that the market may continue to broaden beyond these core customers. Ultimately, Nvidia's success will depend on their ability to continue delivering innovative products and maintaining strong demand signals from their customer base.
NVIDIA's production issues: Despite yield improvements, NVIDIA faced production issues with its Blackwell product, but demand exceeds supply and revenues are expected to ramp up significantly in Q4
NVIDIA's earnings call was unique due to the impact of the semiconductor industry leader on the technology market as a whole. NVIDIA's success in data center GPU fulfillment has been significant, but its impact should not be extrapolated to the rest of the technology sector. The call revealed that NVIDIA experienced production issues with its Blackwell product, but yield improvements are underway, and volume production will begin in Q4. Demand for Blackwell exceeds supply, and revenues are expected to ramp up significantly. Beyond hyperscalers, which represent 45% of NVIDIA's data center business, there is strong demand for accelerated computing across various industries and applications, including generative AI, database processing, transcoding, scientific simulations, and computer graphics. NVIDIA's dominance in the data center market may be causing some companies to buy less from NVIDIA in the near term, but the overall demand for accelerated computing remains robust.
Sovereign AI: Governments and regional service providers are investing heavily in Sovereign AI infrastructure, recognizing digital knowledge as a valuable national resource, and aiming to increase performance and efficiency with new technologies.
Sovereign AI, which refers to AI infrastructure built and funded by specific governments or regional service providers, is becoming a significant global trend. With billions of dollars being invested this fiscal year, countries recognize their digital knowledge as a valuable national resource, requiring harvesting, processing, and transformation into digital intelligence. This trend is expected to continue as more countries build out their AI infrastructure. Additionally, the energy requirements for next-generation AI models will increase due to their greater compute needs. Nvidia aims to address this by increasing performance and efficiency with new technologies like Hopper and liquid cooling. Despite the current demand for AI products outpacing supply, Nvidia expects to see improvements in supply conditions throughout the next year.
NVIDIA's focus on GPU Cloud optimization: NVIDIA optimizes its GPU Cloud within major cloud providers to ensure top performance and TCO, while also offering AI model creation services as an 'AI foundry' for other companies, benefiting from the growing demand for advanced AI technologies.
Foundation model makers, like NVIDIA, are seeing significant growth as these models expand in size, learn various languages or modalities, and are adopted by an increasing number of industries and entities, from startups to enterprises and different countries. NVIDIA, specifically, is not planning to become a cloud compute provider itself but is instead focusing on optimizing its NVIDIA GPU Cloud within other major cloud providers to ensure the best possible performance and TCO for its users. NVIDIA uses its own GPU Cloud extensively for its own AI needs and also offers AI model creation services as an "AI foundry" for other companies. This strategy benefits NVIDIA by ensuring the best version of its technology is available across various clouds, meeting its own needs, and helping other companies. The demand for these advanced AI technologies is growing, leading to substantial expansion in this field.