Following the announcement that NVIDIA and Intel will jointly develop data center and PC chips as part of an investment agreement, Jensen Huang revealed during Thursday's press conference call that he had private discussions with Intel CEO Pat Gelsinger regarding this collaboration, describing Gelsinger as an "old friend." Gelsinger confirmed during the meeting that he and Huang have known each other for 30 years.

"We believe this will be an extremely valuable investment," Huang stated.

NVIDIA announced it will collaborate with Intel to create AI systems for data centers that combine Intel's x86-based central processing units (CPUs) with NVIDIA's graphics processing units (GPUs) and networking technology.

Additionally, Intel will launch CPUs integrated with NVIDIA GPUs specifically designed for PCs and laptops.

Greg Ernst, head of Intel's revenue division, wrote in a LinkedIn post that the deal took months to finalize, with both parties reaching agreement on Saturday.

This investment highlights how the industry positions of these two Silicon Valley companies have reversed since OpenAI launched ChatGPT in late 2022, triggering the generative AI boom.

Over the past five years, Intel's stock has declined 31.78%, while NVIDIA's shares have surged 1,348% as of Thursday's opening. NVIDIA currently maintains a market capitalization exceeding $4.25 trillion, while Intel's market cap stands at only $143 billion.

NVIDIA's AI systems currently utilize Arm-based CPUs rather than Intel's x86-based processors. Huang indicated during the conference call that NVIDIA will soon support Intel CPUs in its AI-focused NVLink racks.

"We will purchase these CPUs from Intel and integrate them into 'super chips' to form our compute nodes, ultimately incorporating them into rack-mounted AI supercomputers," Huang explained.

Huang noted that NVIDIA will also provide GPU technology for Intel's laptop and PC chips—a market that remains underdeveloped. He stated that the combined potential market size for these two product collaborations totals $50 billion.

"We will become a significant customer for Intel CPUs while also serving as the primary supplier of GPU chiplets needed for Intel's processors," he said.

Huang emphasized that this deal with Intel "will not" impact NVIDIA's business relationship with Arm.