Intel CEO Pat Gelsinger Fires AI Shots At NVIDIA, Wants To Crush The CUDA Market
Gelsinger believes the future of AI will hinge on inference technology instead of training, saying that "fundamentally, the inference market is where the game will be at.” A byproduct of this shift in focus from training to inference is there will be less of a reliance on NVIDIA’s CUDA technology, making it easier for companies such as Intel to make some headway in the AI market.
Another reason for Gelsinger’s bullishness on the end of NVIDIA being the dominant player is the desire for the industry to rely on more open standards.
Not puling any punches, Gelsinger commended, “We think of the CUDA moat as shallow and small, because the industry is motivated to bring a broader set of technologies for broad training, innovation, data science, et cetera."
He also pointed to the efforts put in place by Google and OpenAI to make AI training more open by using the Python programming language. As such, Gelsinger believes "the entire industry is motivated to eliminate the CUDA market." Strong words for sure.
The future Intel envisions is one where a user has access to both computing resources on their PC working alongside resources available in the cloud. It’s a future where Intel can be successful and make a lot of money. However, it needs Gelsinger to be correct on his no-holds barred thoughts regarding NVIDIA’s decline in the AI space, but at the moment that seems like it will be an uphill battle.
