The AI boom enters its post-Jensen era

• But replacing a CEO willing to bankroll startups, backstop billions in cloud debt and aggressively fund rivals to the AI giants will be much harder.

The big picture: The next chapter without Huang will be more cautious, but the industry is capable of absorbing the change, for two key reasons:

1. New chips - Semiconductor competition has become harder to contain.

• At OpenAI and other startups, a longstanding joke was that the best way to get a discount on Nvidia chips was to openly talk to Google or others about alternatives. Huang would often swoop in enthusiastically, eager to lock in a customer.
• The progress of custom chips from Google, Amazon, Microsoft and OpenAI, as well as new deals with Broadcom and Nvidia rival AMD, gives those conversations more substance. Customers have options, and far more leverage than ever.

• AI itself has also made those offerings easier to use. Anthropic says its Mythos model helped run certain workloads on AMD chips in about three days, compared with six to 12 months in the past.

2. Washington - In DC, China hawks are ascendant. Huang's departure strengthens calls from administration forces, some lawmakers, and companies such as Anthropic that have pushed for tighter controls on chip sales.

• Legislation is rapidly advancing to cut off loopholes and strengthen export restrictions for advanced chip sales to China, and the Commerce Department is reportedly preparing additional ways to limit such access.
• Cutting off those routes could help protect America's lead and offer a major pre-IPO boost to Anthropic and OpenAI.

Reality check: Alternative chips and leaps forward in software won't solve the near-term computing crunch that all the big players in AI are facing.

• Nvidia's chips remain the most sought-after because they are the most widely supported, making them critical for the industry's smaller players and important for the big companies who need maximum efficiency out of each data center they build.

There are two reasons a strong Nvidia still remains essential:

1. Finance - The company's checkbook will be far more difficult to replicate than its computing power.

• Nvidia has made $530 billion of off-balance-sheet commitments and guarantees, including to buy memory and backstops for data centers and certain companies building them, according to data firm SemiAnalysis.
• Smaller cloud providers, often called neoclouds, as well as a vast array of startups all over the world, have lost perhaps their most enthusiastic champion.
• Nvidia will remain a formidable competitor, but it will almost certainly become less aggressive about startup funding, acquisitions and backstops, tempering its efforts to spread exposure among institutional investors in ways that some critics said resembled the mortgage bubble.

2. Open models - Nvidia has been the most enthusiastic funder and champion of the Western open-model ecosystem—giving developers cheaper alternatives to the expensive offerings from frontier AI companies.

• Its pending acquisition of Hugging Face, licensing agreement with model-developer Poolside, backing of Reflection and development of the Nemotron family of open models extend well beyond chip sales.
• Without Nvidia, it's not clear whether the U.S. will be able to develop open-weight alternatives to China, since many don't have a clear or easy path to financial returns given that such systems are often free to download.

What we're watching: Whether the industry's growing freedom to choose its chips can compensate for an Nvidia that is suddenly much less willing to finance its ambitions.

Bottom line: Huang's departure won't derail America's AI ambitions. But it leaves a hole in the ecosystem that will be harder to fill than any shortage of Nvidia chips.