Google Cloud CEO Thomas Kurian has publicly pushed back against the narrative of a “chip war”, wherein custom AI chips like Google’s Tensor Processing Units (TPUs) are seen as a threat to Nvidia’s dominance. Kurian’s comments have come about a month after a turbulent week where news of Facebook-parent company partnering with Google to train its Llama AI models on search giant’s in-house silicon wiped billions from Nvidia’s market value.
Google-Nvidia not a rivalry: Google Cloud chief Thomas Kurian
Kurian characterised the relationship between Google and Nvidia not as a rivalry, but as a partnership, dismissing the idea that the rise of custom chips is a “zero-sum game”, as per a report by Fortune. “For those of us who have been working on AI infrastructure, there’s many different kinds of chips and systems that are optimised for many different kinds of models,” Kurian said. He noted that Google continues to optimise its Gemini models for Nvidia GPUs, even allowing them to run on Nvidia clusters.
Nvidia’s response on Google-Meta partnership
Nvidia pushed back against the narrative of Google having better chips, insisting that its GPUs remain “a generation ahead of the industry” even as Google’s in-house AI chips gain momentum. The development fuelled concern among investors and analysts, which led to a 3% drop in Nvidia’s shares.“We’re delighted by Google’s success — they’ve made great advances in AI and we continue to supply to Google,” the company said in a post after Google-Meta partnership news broke.“NVIDIA is a generation ahead of the industry — it’s the only platform that runs every AI model and does it everywhere computing is done. NVIDIA offers greater performance, versatility, and fungibility than ASICs, which are designed for specific AI frameworks or functions,” the company added.Notably, while Nvidia still commands over 90% of the AI chip market, hyperscalers like Google and Meta are increasingly looking at “mixed chip infrastructures” to balance raw performance with cost-effectiveness.
Leave a Reply