Nvidia and AMD takeaways from CES and the next stage of AI

Key Concepts

• Physical AI: AI systems that understand and interact with the physical world, going beyond purely digital applications.
• GPU (Graphics Processing Unit): Specialized electronic circuits designed to rapidly manipulate and display computer graphics, now crucial for AI computation.
• TAM (Total Addressable Market): The total market demand for a product or service.
• High Bandwidth Memory (HBM): A high-performance RAM interface for 3D-stacked synchronous dynamic random-access memory (SDRAM).
• Chiplets: Small functional blocks of integrated circuits that can be combined to create larger, more complex chips.
• AI Supercycle: A period of rapid growth and innovation in the artificial intelligence sector, driving demand for related technologies.
• Reuben Chip: Nvidia’s next-generation GPU architecture, expected in the second half of 2026.
• Yataflop: A unit of measurement for computing performance, specifically used by AMD to highlight their chip capabilities.

The Nvidia Dominance & AI Landscape Following CES 2024

The discussion centers around Nvidia’s impactful presence at CES 2024, led by CEO Jensen Huang, and the broader implications for the AI industry. The core argument is that the AI revolution is not a bubble, but a fundamental shift in the computing landscape, with Nvidia positioned as a dominant force.

Nvidia’s CES Performance & “Physical AI”

Jensen Huang’s keynote, highlighted by his signature black leather jacket (specifically an alligator leather jacket sparking online speculation), served as a signal of Nvidia’s continued innovation and market leadership. The key takeaway from the presentation was the emergence of Physical AI – AI systems capable of understanding and interacting with the physical world. Huang emphasized that AI is expanding beyond digital applications into areas like autonomous vehicles and robotics, stating, “Wherever the universe has information, wherever the universe has structure, we could teach a large language model…to turn that into an AI.” This represents a significant evolution, moving towards AIs that understand “the laws of nature.”

The Shifting Computing Paradigm

Huang articulated that the computer industry undergoes a reset every 10-15 years, and currently, two simultaneous platform shifts are occurring: the rise of AI and a fundamental reinvention of the entire software stack. He stated, “Applications are now going to be built on top of AI…how you run the software, how you develop the software fundamentally changed. The entire five stack of the computer industry is being reinvented.” This signifies that AI is not merely an application within the existing computing framework, but a catalyst for a complete overhaul.

Demand & Market Valuation

The demand for AI chips, particularly Nvidia GPUs, is experiencing exponential growth. Models are increasing in complexity by a factor of 10 annually, driving a corresponding surge in computational requirements. This strong demand is pushing back against concerns about an AI bubble and suggests that earnings estimates for the S&P 500 may be significantly understated (potentially by 10-20%). Despite this, the panel expressed bewilderment at Nvidia’s stock trading at a multiple inline with the S&P 500, suggesting it is undervalued given its market position. Dan Ives stated, “Nvidia’s stock…almost makes no sense to me.”

Roundtable Discussion: Perspectives on Nvidia & Competitors

The “Opening Bid” roundtable featured Dan Ives (Wedbush Securities), Gil Luria (DA Davidson), and Enzy Yao (Yahoo Finance) offering their insights.

Dan Ives: The “Godfather of AI” & a Tech Bull Market

Ives characterized Jensen Huang as the “godfather of AI” and dismissed the notion of an AI bubble, emphasizing the trillions of dollars being invested in the space. He highlighted Nvidia’s blueprint for physical AI and positioned the current market as a “tech bull market” with significant upside potential. He also praised AMD, stating, “If Lisa Sue is flying the plane, I'm in 3A feeling really good watching Netflix.”

Gil Luria: Data Centers & the Edge

Luria focused on Nvidia’s success in creating a massive data center market and its strategic shift towards deploying GPU content “at the edge” – on devices like automobiles and robots. He questioned the timing of this transition and its ability to sustain growth as the data center cycle potentially peaks. He also emphasized the importance of software companies like Snowflake and Datadog in enabling enterprise AI adoption.

Enzy Yao: Advanced Packaging & Intel’s Potential

Yao highlighted Intel’s advanced packaging technology (chiplets) as a potential area of strength, noting Jensen Huang’s investment in this area. She pointed out that this technology allows Intel to stack and combine smaller chips, potentially offering a competitive advantage.

Micron & the Memory Supercycle

The discussion shifted to Micron, a key player in the memory market, particularly High Bandwidth Memory (HBM) crucial for AI servers.

• HBM Demand: The Total Addressable Market (TAM) for HBM is projected to reach $100 billion by 2028, growing at a 40% compounded annual growth rate.
• Valuation Discrepancy: Micron trades at a P/E ratio of 9.9, significantly lower than the S&P 500 (22) and Nvidia (25), despite strong catalysts.
• Memory Bottleneck: The panel agreed that a bottleneck in memory supply is driving up prices and benefiting memory manufacturers.
• Competitive Landscape: SK Hynix was identified as a leading competitor in the memory market, with Dan Ives describing a Micron stock purchase as “like buying a Mickey Mantle signed car at a garage sale.”
• Commodity Risk: Gil Luria cautioned that memory remains a commodity, with limited pricing power due to interchangeability between suppliers.

Profiting from the AI Revolution: Nvidia, AMD, & Intel

The final segment addressed the best strategy for investing in the AI revolution.

• Nvidia as the Leader: The consensus favored Nvidia as the most reliable investment, given its market dominance and continued innovation.
• AMD as a Potential Challenger: AMD was seen as a viable option, with potential to gain market share.
• Intel as a Speculative Play: Intel was considered a higher-risk, higher-reward investment, contingent on its turnaround efforts and success in advanced packaging. Dan Ives suggested including Intel in a portfolio, noting it’s “no longer at that table by the sort of kitchen.”

Conclusion

The discussion paints a picture of a rapidly evolving AI landscape, with Nvidia firmly established as the leader. The emergence of Physical AI, the reinvention of the computing stack, and the exponential growth in demand for AI chips are driving a significant investment opportunity. While AMD and Intel offer potential alternatives, Nvidia’s current position and continued innovation make it the most compelling investment for capitalizing on the AI revolution. The memory market, particularly HBM, is also poised for substantial growth, with Micron and SK Hynix as key players. The overall sentiment is optimistic, with the panel dismissing concerns about an AI bubble and emphasizing the long-term potential of this transformative technology.