How Anthropic and Google TPUs Are Challenging NVIDIA’s AI Chips Dominance

Anthropic is reassessing the chip market while also considering the general usage of NVIDIA graphics processing units that underpin its high-performance computing systems being developed, and its future GPUs will be executed. Rising demand for computing underpins this trend as well, making it standard for companies in general. Anthropic already reserved 3.5 gigawatts of Google TPUs focused computing Power till 2027 using its recent agreement with Broadcom.

 The company has signed a multi-year lease and demonstrated its plan to diversify usage among Google TPU, AWS Trainium and NVIDA GPU on their quarterly issued AI investor presentations.AI semiconductors is the largest business unit within AI cluster for Broadcom. In Q1 FY2026 Q1-FY2027 Approximately, Broadcom had reported $8.4 Billion in revenues from aforementioned segment which is 106% more than their original guidance. So, henceforth the organizations are renegotiating their equipment policies for gaining an edge in technology. 

According to the experts in the field of technology NVIDIA will become one among those few manufacturers of chips which produces advanced chips. The software platform developed by NVIDIA called CUDA is being utilized by many developers .This is due to greater degree as compared to other chip manufacturers high obstacles set on system training product entry on markets. Anthropic has intentions of broadening its scope from just using NVIDIA despite its plans however, CNBC reported that it is actually contemplating the construction a rent synonomous with GPUs of using NVIDIA systems. 

Big Tech like Google.com, Amazon and Microsoft are increasingly designing and making their own custom chips to reduce their reliance on external suppliers. Concerns about access to certain types of components, particularly those essential for fabricating chips, as well as the expense of computing, will continue that trend, forcing firms to design more of their own infrastructure. But some tech companies, in their pursuit of flexibility, also see NVIDIA as key. Its GPUs are fast and can be used with other tools.

Analysts expect there will indeed be a gradual shift, as many big tech companies will also try and find a way to use NVIDIA’s hardware with which to do work on other processors and in other clouds. But this too will take time; perhaps longer, given the performance of NVIDIA’s turnkey offering. The company is currently not just the market leader, it has kept pace with all threats to its revenue supremacy.

Get In Touch with TheTweaks for daily Update.