Google wants to build an alternative to Nvidia’s empire
The group supports the construction of data centres and cloud platforms to create an alternative ecosystem based on TPUs
Is Nvidia’s leadership in artificial intelligence chips at risk? Not at the moment, but if players such as Google decide to enter the fray, things could change.
Big G is, in fact, reportedly working to transform its AI chips from in-house technology into a new line of business, following a strategy that closely mirrors the one adopted by Nvidia in recent years.
According to the Wall Street Journal, the Mountain View-based group is using its financial clout to support the development of AI infrastructure based on Tensor Processing Units (TPUs) – processors developed in-house and used for several years now to power services such as Search, YouTube and Gemini. The aim seems quite clear: to reduce the market’s dependence on Nvidia processors and create an alternative ecosystem capable of attracting customers and investors. After all, there is a great deal of money at stake.
According to sources, however, the real winning factor could lie in something that goes beyond the ‘mere’ chip. According to the Wall Street Journal, the Mountain View giant is reportedly seeking to replicate one of the strategies that contributed to Nvidia’s success: using its financial clout to encourage the construction of data centres based on its own technologies.
In practice, Google would not merely develop and sell processors for artificial intelligence, but would also (at least in some cases) help to facilitate the investment needed to build the infrastructure that will utilise those chips.
