Nvidia's AI Power Play

Good morning. Few people can predict AI's impacts and opportunities well. However, Nvidia CEO Jenson Huang has a decent track record. When he talks strategy, I listen. And there are always some nuggets I take away.

There are hundreds of podcasts and essays about Nvidia's strategy, and I highly recommend the one on Acquired.

But these interviews are coming of age so fast with the rapid development of AI that it’s always worth another look, especially after investor presentations.

Today, I am reviewing the main takeaways from Nvidia's earnings event last week and their recent investment strategy. As usual, I am looking between the lines and finding lessons for our strategy muscles + I will include the bombshell news that came in while editing this post—escalating antitrust investigations on Nvidia’s AI dominance.

Nvidia’s recent investments

Over the last year, Nvidia invested in more than 30 startups. That’s 3x from the previous year.

By the way these investments are up $1.25 billion (from $300 million to $1.55 billion as of Jan ‘24).

Jensen Huang is known to look at the potential of investments long in the future. It will not be two years in the future, but it will be more like ten years out. Here is a list of recent bets Nvidia made with high expectations:

• Inflection AI ($1.3 billion) - Creating a personalized AI assistant

• Databricks ($500 million) - Data, Analytics and AI

• Recursion Pharmaceuticals ($50 million) - A drug-discovery company

• Cohere - Builds generative AI models

• SoundHound AI - Voice-enabled artificial intelligence

• Moon Surgical - Develops chips in the life-science field

• CoreWeave - Operator of data centers

Notice that these companies are well positioned along the whole value chain of artificial intelligence.

Nvidia likely investing in OpenAI

Two recent AI bombshell announcements:

1. OpenAI is setting to raise a record round at a $100 billion valuation

2. Nvidia and Apple are in talks to invest in OpenAI

Nvidia is continuing its multi-edge AI investment strategy.

With these investments, Nvidia is at the forefront of AI research and development, where it interfaces directly with users. The challenges of companies like OpenAI can inform Nvidia's further chip design decisions and let Nvidia tailor its products more towards OpenAI and get the upper hand in being selected as a supplier.

On the defensive side, Nvidia wants to stay the main partner for OpenAI’s ambitious endeavors. And of course, being an investor also gives them dips on being their preferred supplier.

Companies deploy venture capital arms to stay ahead of the competition. The AI industry is one of the fastest-moving technologies. Venture capital investors typically have access to more information when they invest in private companies.

Nvidia benefits at multiple points from this investment:

• Investor upside if OpenAI grows beyond a $100 billion valuation… crazy, I know

• OpenAI works closely with Nvidia on their next AI technologies - chips are tailored toward OpenAI’s needs

• Learn about relevant end-user applications and be at the forefront of AI research and development

Apple is also planning to invest. Apple is on the other side as a customer of OpenAI’s services. They announced their partnership at WWDC in June this year.

These are very strategic investors for OpenAI. Having the main supplier of scarce AI hardware as an investor could suggest a high rank on Nvidia’s customer waitlist for more processing power. Having Apple as a customer and investor should help their long-term partnership and maybe deeper integrations in Apple’s ecosystem.

Nvidia’s vision of an AI factory

Last week, at the earnings report, Jensen Huang spilled more beans on the company's future focus. This completes the puzzle. The master plan is to offer software, data center design services, and networking technology in addition to its current products.

In addition, Nvidia also started offering the Grace Superchip (CPU). This is another product that extends beyond their current GPU offering. Based on Nvidia, it is optimized for their market-dominating GPUs and is sold together. It sounds like a bundled opportunity, doesn’t it?

With this, Nvidia will become a one-stop shop for all things AI model development infrastructure or, as Jensen Huang calls it, “AI factories.”

And although a completely bundled “AI factory” is a wonderful product for interested business customers to “plug and play,” it does raise antitrust concerns.

We discussed this two weeks ago when we did our deep dive into Google Search being bundled into Safari and Android devices.

That Escalated Quickly.

While editing this post, the news broke that the DOJ had indeed sent subpoenas to Nvidia. With that, the Nvidia antitrust investigation is escalating quickly.

The investigation was launched in August this year:

Nvidia has invested in this area for almost twenty years when competitors were twiddling their thumbs. Now, they are earning the dividends.

Nvidia GPUs are rightfully leading the market because they are the best product.

However, the concerns are related to offering product bundles that could leave the competition no chance to compete. Nvidia has a rare but much-needed product. Nobody comes close to their GPU performance, and everyone wants it. Offering bundle deals could lead to manipulating customers into selling product extensions.

If you were in the market for a PlayStation 5 when they came out during the COVID chip shortage, you know exactly what I am talking about. You couldn’t get your hands on a PS5 unless you committed to a $1,500 bundle with crap you didn’t need.

Key Takeaways

1. A bold investment strategy: Nvidia has tripled its venture investments over the past year. Its investments in 30 startups span data, healthcare, robotics, and genAI and position Nvidia along the entire AI value chain.

2. Future-proof investment in OpenAI: OpenAI is seen as one of the few leaders in genAI. Nvidia wants to be close to the customer-facing side, research, and development to understand exactly what needs to be done to drive its products in the right direction.

3. AI Factories: Nvidia wants to become a full-stack provider for anyone who wants to create artificial intelligence models. It invests heavily in its expansion to fill the current gaps in its product portfolio.

4. Antitrust concerns: Dominating 80% of any market shows a superior product performance. But bundling products and getting drawn into dirty pricing games could get them in trouble.

I want to highlight Nvidia’s aggressive positioning in the AI sector. It shows the proper foresight in navigating future trends and shows how they want to be in a position to build everything that an AI data center needs across several industries. However, some threads emerge from regulation. They need to tread lightly to navigate their future expansion wisely.

Tech-Nostalgia: Interesting facts about Youtube

Jawed Karim’s (co-founder of YouTube) first email describing the idea of YouTube recently went viral. For nostalgic reasons, I’ll take this opportunity to collect a few interesting facts about YouTube.

Here is the email:

The evolution of the video playbar over the years

I hope you enjoyed this episode. I love your reactions to my writing and look forward to your weekly comments.

As usual, sharing with other interested readers helps our community grow and is much appreciated. 🙏

Have a great rest of the week,