On October 28, 2026, the tech landscape shifted as reports emerged that OpenAI is reportedly planning to purchase Google's AI chips. This move aims to meet the growing demand for AI computing and would deepen the relationship between the two tech companies. The Information reported on this, citing a source who indicated that the company is exploring the purchase of AI chips from Google to bolster its AI infrastructure.
The backdrop to this announcement is Nvidia's market-leading status in AI chips. For years, Nvidia has held the crown, but the tension is palpable among competitors. Can Nvidia lose its dominance? This question hangs over the industry as other major players seek alternatives. Google's Tensor Processing Units, or TPUs, have gained traction as an alternative to Nvidia's market-leading AI chips. Google began selling TPUs directly to some customers this year, signaling a shift in how cloud providers distribute hardware power. Investors will be watching closely, as Alphabet will report its quarterly results later, with investors looking for signs that its heavy investment in AI is driving growth in its cloud business.
While OpenAI considers buying Google hardware, another giant in the space, Anthropic, is taking a different approach. Anthropic is diversifying its AI chip platforms across four chip families. This strategy could reduce its dependence on Nvidia and give it more negotiating leverage. Ashish Nadkarni noted regarding this approach: "A multichip approach also makes Anthropic less vulnerable to fluctuations in Nvidia’s chip supply." By spreading its investments across a diverse range of compute options, Anthropic seeks to mitigate risks associated with a single supplier. This contrasts with OpenAI's direct procurement interest, yet both reflect a broader industry desire to avoid reliance on one vendor.
He stated: "Other major companies, including OpenAI, likely use a multichip approach.
In parallel with these moves, Oracle will also offer TPUs through its cloud-computing infrastructure. This expansion adds to the competition in the AI computing market. As Google begins selling TPUs directly to some customers this year, the ecosystem is becoming more fragmented. Patrick Moorhead commented on the shifting dynamics: "The real exposure could be Nvidia’s pricing power."
Gil Luria provided insight into the broader adoption of this strategy. He stated: "Other major companies, including OpenAI, likely use a multichip approach." This implies that the decision by OpenAI to consider Google chips aligns with industry norms rather than being an isolated anomaly. However, not everyone believes the shift will completely neutralize Nvidia's advantages. Brad Gastwirth offered a critical perspective: "Having more chip options does not eliminate Nvidia’s advantages in performance, networking and software." It involves balancing cost, availability, and raw capability.
The stakes involve billions spent on Nvidia computing capacity, particularly by Anthropic. As OpenAI explores purchasing Google's AI chips, the goal is to secure sufficient resources without compromising on the quality needed for advanced AI work. The diversity of compute available is crucial for scaling these operations.
As of October 28, 2026, the situation remains fluid. OpenAI plans to buy Google's AI chips, though the final details are still unfolding. Anthropic continues to diversify its AI chip platforms across four chip families. Google maintains its position by offering TPUs to some customers directly. The industry is watching to see if Nvidia can maintain its lead despite these challenges. The narrative points toward a future where no single company controls the entire stack of AI computing infrastructure. OpenAI is exploring the purchase of AI chips from Google to bolster its AI infrastructure, competing with Nvidia's dominance. Ultimately, the focus remains on meeting the growing demand for AI computing while managing the risks of dependency.
