TLDR;
Beth Kindig from the IO Fund discusses Nvidia's financial results, the AI landscape, and investment strategies. She emphasizes that Nvidia is evolving into a rack-scale systems company, not just a chip company, and highlights the importance of networking and energy in future GPU generations. Kindig also touches on the potential overvaluation in AI software, opportunities in AI networking and energy sectors, and a risk-managed approach to crypto investments, advocating for profit-taking in volatile markets.
- Nvidia is transitioning into a rack-scale systems company.
- AI networking and energy are key areas for future investment.
- AI software may be overvalued, while hardware remains critical.
- A risk-managed approach to crypto is essential, with profit-taking strategies.
Nvidia’s Latest Financial Results: What’s the Real Takeaway? [1:07]
Beth Kindig analyzes Nvidia's recent earnings, noting a slight decline in compute in Q2 due to China, but emphasizes that the forward look for Q3 saved the day. The focus should be on GPU generations rather than quarterly earnings, with the Blackwell generation expected to significantly impact earnings starting in Q3. A 15% quarter-over-quarter acceleration in data center is anticipated, with networking continuing to explode. The strength of Q3 indicates that the company is shipping new systems and that there are green lights across the board.
China GPU Clampdown: Headwind or Background Noise? [3:46]
Kindig believes the impact of China on Nvidia's revenue is overblown. In the last quarter, China accounted for a $4 billion impact, while the Blackwell generation is projected to generate roughly $28 billion. Losing China revenue now is better than during the trough when Hopper sales were cooling off. The numbers support that Blackwell is shipping, and the 15% quarter-over-quarter acceleration indicates that China's absence doesn't impact growth in the second half of the year.
Could China Flip from Zero to Surprise Upside? [6:08]
China will always be an opportunity for Nvidia, but Kindig believes China aims for self-reliance in technology. Given that AI is a larger market than e-commerce, social media, EVs, and mobile, China is unlikely to depend on United States silicon. Onshoring supply chains is a significant undertaking that could take a decade, with global tensions being the biggest question mark for the semiconductor industry.
Valuation vs Story: Is NVDA Still a Buy? [8:31]
Nvidia is no longer just a chip company; it's a rack-scale systems company with software and networking components. This makes the stock investable for the foreseeable future. Nvidia's parallel processing power and tensor cores give it an insurmountable lead in AI training. Those deeply entrenched in the hardware stack, like Nvidia, will also be dominant players in software, similar to Apple's dominance with the iPhone and iOS.
Tactics: Add Now, Wait for Pullbacks, or Trim? [12:47]
The most important question is whether Nvidia, as the world's most valuable company, can continue to grow. Kindig's firm conducts extensive research on Nvidia, using a 100-point checklist multiple times a year. The data center market is currently at a $200 billion run rate, and analyst estimates are too low. By the end of next calendar year, it's expected to reach $75 billion per quarter, creating room in the valuation. Blackwell and Blackwell Ultra should bring the data center to $300 billion, and Kindig believes Nvidia will hit $500 billion by 2028. This growth is driven by big tech companies investing heavily to avoid being left behind.
What Could Actually Cool Nvidia Sales? [18:30]
Nvidia's product roadmap is aggressive, which can lead to thermal and power management issues, as seen with the six-month delay in the Blackwell generation. Additionally, energy could become a bottleneck as future GPU generations require significantly more power. The grid may be maxed out by the Reuben generation, which could require 3 to 5x more power than current systems.
Tech Bubble Check: Who Wins, and Who Gets Left Behind? [21:27]
AI software is currently in the R&D stage, and the bubble is primarily at the software layer until real revenue growth is seen. Picks and shovels like Nvidia and its suppliers are benefiting from massive capex spend, while others are still trying to find their place. It's crucial to distinguish between those participating now and those just trying to fit in. AI is an enterprise technology first and foremost, allowing companies to optimize internal processes and products.
Tech Moats: Is AI Spend Defensible? [27:52]
Big tech companies are building fortresses with AI, making their ad engines superior and enabling continued growth. AI is also a threat to their moat, particularly regarding which company can become the foundation model for developers. OpenAI has done well in this area with its API usage.
Beyond NVDA: Where Beth Sees the Next Edge [30:35]
AI networking is where Kindig's portfolio has an outsized allocation. As Nvidia moves from eight GPUs communicating to 72 GPUs, it creates a networking problem to solve. This requires 5x to 9x more networking components, such as retimers and fabric switches, to ensure GPUs can communicate without signal loss.
Power Problem: Is Energy the Next Big Trade? [33:18]
Kindig is in the early stages of moving into energy investments. She models at least 10x growth for the right energy stocks between Blackwell and 2030, as energy requirements for GPUs are rapidly increasing. Big tech companies are gobbling up alternative power sources to prepare for future GPU systems.
⭐️Wealthion Golden Nugget: Beth’s Crypto Trade Plan and Bitcoin Outlook [36:12]
Kindig's company uses a different process for Bitcoin and crypto, viewing it as sentiment-driven with no earnings reports or financials. Technicals work well for risk management in crypto, and they advocate setting a price target years in advance and trimming when it's hit to buy lower. A near-term top could be seen in the 130s, where they will take profits and put them back to work lower. Their crypto mix is usually 70% stocks and 30% crypto, and they believe in booking profits, particularly in a volatile asset class like crypto.