NVIDIA’s OpenAI partnership made headlines in the past few weeks, considering how monumental the deal was, but there’s a background context to it that quite a few are aware of.

NVIDIA’s CEO Found It an ‘Absolute Priority’ To Strike a Deal With OpenAI, To Ensure Its Dominance in The AI Segment

For those unaware, NVIDIA had recently entered into a ‘mega deal’ with OpenAI, where Jensen & Co. would invest $100 billion into the non-profit AI firm, and in turn, the AI giant would get ‘millions’ of AI chips, worth ten gigawatts of power. Now, in a report by the WSJ, it is revealed that the NVIDIA-OpenAI deal was stalled during the initial talks back in the summer, but once OpenAI was rumored to integrate Google TPUs for AI workloads, NVIDIA’s CEO rushed back to Sam Altman to secure the stalled deal. NVIDIA’s direct investment was aimed at helping Altman get the required compute power, while keeping the firm’s chips in the spotlight.

That spring, OpenAI also signed its first cloud contract with Google, raising the specter that it might begin using its chips as well.

After the story was published, Huang called Altman asking him if it was true, and signaled that he was open to getting the talks back on track, according to people familiar with the matter. Huang realized that Nvidia could help OpenAI by making a direct investment in the company, a person who spoke to him said.
– WSJ

Well, Jensen was paying attention to the headlines around his company, it seems, and more importantly, the sudden change in stance clearly indicated that securing a deal with OpenAI was a top priority at that time, which is why the collaboration with the firm is one of NVIDIA’s biggest deals yet. Since OpenAI is one of the biggest AI firms out there, any collaboration involving Google’s custom AI chips means that NVIDIA’s tech stack does have an alternative out there, and this certainly isn’t optimal for Team Green to maintain its AI dominance.

More importantly, the $100 billion investment acts as a ‘supplier lock-in’ for NVIDIA in this case, since now, OpenAI would be reliant on the firm’s hardware, even up to Vera Rubin systems, which means that for the next few years, Sam Altman is entirely on NVIDIA. So Jensen basically secured two objectives with this deal: the first is that it becomes one of the largest compute providers for OpenAI, and secondly, it does prevent the external adoption of Google’s ASICs, since the chips are known to be a formidable rival to Team Green’s hardware.

It’s still interesting to see how Big Tech deals with itself through a matter of headlines, and for NVIDIA, acting swiftly on such reports is necessary, since the firm needs to maintain its ‘dominant status’ in the AI industry.

https://wccftech.com/nvidia-ceo-scrambled-to-rescue-a-stalled-deal-from-collapse-with-openai/amp/