Jensen Huang is gone, and Nvidia’s customers finally dare to consider buying chips from someone else.
The Information reported that an executive at a cloud provider that buys Nvidia chips revealed the company is seriously considering switching to Google and AMD.
The most interesting part is what came next: they’d wanted to diversify their purchases for a long time, but never dared, because they were afraid of offending Jensen Huang.
There’s a bit of dark humor in that. In the mighty AI computing market, with deals worth tens or even hundreds of billions of dollars, customers choosing a supplier still have to consider how a CEO might feel.
But that’s precisely what shows how the power Huang built over the past few years has long extended beyond the chips themselves.
Nvidia doesn’t just control the most sought-after GPUs. Through investments, credit guarantees, and purchase commitments, it has also become an important source of financial support for many AI cloud providers. It sells the shovels, helps its customers finance the purchase of shovels, and sometimes even rents the holes its customers dig.
In a relationship like this, purchasing decisions aren’t just about value for money. Of course, you can use AMD or consider Google’s TPUs—but what if you lose Nvidia’s support as a result?
A technological monopoly is powerful enough. More powerful still is making customers feel: I’d better not cross them.
More ironically, that same executive is now starting to worry that Nvidia’s chip development over the next few generations could suffer after Huang’s departure.
Before, they were afraid to switch. Now, they’re afraid not to.
Of course, this doesn’t mean Nvidia’s customers will all defect overnight. The CUDA ecosystem, chip performance, and the scale of existing deployments remain real competitive barriers. At least in the short term, Nvidia is hardly short of buyers.
But I think this episode reveals a more interesting question than market share:
Over the past 33 years, was Jensen Huang building a great company that could stand on its own, or a business empire that everyone had to orbit around him?
While he was in charge, the two looked almost indistinguishable.
It wasn’t until he left that customers dared to take out a quote for another company’s products.
The Information reported that an executive at a cloud provider that buys Nvidia chips revealed the company is seriously considering switching to Google and AMD.
The most interesting part is what came next: they’d wanted to diversify their purchases for a long time, but never dared, because they were afraid of offending Jensen Huang.
There’s a bit of dark humor in that. In the mighty AI computing market, with deals worth tens or even hundreds of billions of dollars, customers choosing a supplier still have to consider how a CEO might feel.
But that’s precisely what shows how the power Huang built over the past few years has long extended beyond the chips themselves.
Nvidia doesn’t just control the most sought-after GPUs. Through investments, credit guarantees, and purchase commitments, it has also become an important source of financial support for many AI cloud providers. It sells the shovels, helps its customers finance the purchase of shovels, and sometimes even rents the holes its customers dig.
In a relationship like this, purchasing decisions aren’t just about value for money. Of course, you can use AMD or consider Google’s TPUs—but what if you lose Nvidia’s support as a result?
A technological monopoly is powerful enough. More powerful still is making customers feel: I’d better not cross them.
More ironically, that same executive is now starting to worry that Nvidia’s chip development over the next few generations could suffer after Huang’s departure.
Before, they were afraid to switch. Now, they’re afraid not to.
Of course, this doesn’t mean Nvidia’s customers will all defect overnight. The CUDA ecosystem, chip performance, and the scale of existing deployments remain real competitive barriers. At least in the short term, Nvidia is hardly short of buyers.
But I think this episode reveals a more interesting question than market share:
Over the past 33 years, was Jensen Huang building a great company that could stand on its own, or a business empire that everyone had to orbit around him?
While he was in charge, the two looked almost indistinguishable.
It wasn’t until he left that customers dared to take out a quote for another company’s products.