Team Blue's former CEO Pat Gelsinger takes a dig at NVIDIA's current approach to AI GPU pricing, claiming that the market will eventually realize this.
Intel's Former CEO Says NVIDIA's AI GPUs Are Too Expensive For AI Inferencing, Claims Jensen Got Lucky With AI
Well, it seems like Gelsinger isn't too fond of NVIDIA's progress in the AI market, and given Intel's performance in this particular segment, it does seem justified. In an Acquired podcast at NVIDIA's GTC 2025 venue, Gelsinger was asked about how NVIDIA's CEO, Jensen Huang, came to realize how big AI actually is, and in response, not only did he say that Jensen got lucky with AI, but also expressed that NVIDIA's current hardware stack is too expensive for AI inferencing workloads.
The CPU was the king of the hill, and I applaud Jensen for his tenacity in just saying, "No, I am not trying to build one of those; I am trying to deliver against the workload starting in graphics." And, then, he got lucky, right with AI.
Today, if we think about the training workload, okay, but you have to give away something much more optimized for inferencing. You know a GPU is way too expensive; I argue it is 10,000 expensive to fully realize what we want to do with the deployment of inferencing for AI and then, of course, what's beyond that.
It is safe to say here that NVIDIA was the one who stayed committed to the "long-ball" game with AI, and that is what Gelsinger has applauded Jensen for, sticking to his instincts. Intel's former CEO had previously expressed concerns about NVIDIA's AI progress, labeling CUDA as a "moat" and that the real future lies in inference, but despite such comments, Team Blue never really did manage to capitalize on the AI hype, earning far less revenue compared to what its competitors were able to.
As far as where Intel stands right now, the firm has abandoned its much-anticipated "Falcon Shores" AI lineup and has bet its future on next-gen "Jaguar Shores. Moreover, Intel's Gaudi lineup has shown significantly underwhelming performance compared to counterparts like Instinct or Hopper, which makes it clear that Intel hasn't managed to find its footing in the AI market. With Intel's new CEO, Lip-Bu Tan, there's optimism towards the future of the company, but nothing is concrete as of now.
Gelsinger also said that quantum processing is likely to be the next form of computing the industry will focus on in the future, and the technology might see market adoption by the end of this decade.
