There is an old saying circulating in Silicon Valley that during the gold rush, the best money is made by shovel salesmen. But in the age of generative artificial intelligence, this relationship is evolving into something much more complex: a shovel manufacturer is just laying out billions to keep its biggest digger at work.
Nvidia is close to finalising a $30 billion investment in OpenAI. The move is part of a giant funding round that is expected to value the ChatGPT creator at an astronomical $830 billion. While these sums are mind-boggling, for Jensen Huang, it is a strategic hedge of supply chain and demand all in one.
The mechanism of this deal resembles a business perpetual motion machine. OpenAI needs massive computing power to train next-generation models, and Nvidia needs a guarantee that its most expensive H100 and Blackwell systems will have a steady customer. Most of the capital that Nvidia now ‘gives’ to OpenAI will come back to it in the form of processor orders. This is essentially greasing the gears of an ecosystem in which the two companies are interdependent.
This round, which also involves SoftBank and Amazon, sheds light on the new power structure in the technology sector. The boundaries between hardware vendors, cloud giants and software developers are blurring. Nvidia, traditionally associated with component manufacturing, is becoming a key financial architect of the industry, ensuring that its largest customers do not lose momentum in the AI arms race.
The barrier to entry for ‘General Artificial Intelligence’ (AGI) has ceased to be measured in algorithms and has begun in hundreds of billions of dollars and access to silicon. The partnership with OpenAI, which took longer than expected to negotiate, shows that even giants have to tread carefully on a ground full of antitrust regulations and technical challenges. Ultimately, however, at a valuation of $830 billion, OpenAI is becoming too big for Nvidia to let it seek solutions from competitors.
