Intel to Build GPUs as It Pushes Deeper Into AI Hardware
Chief Executive Lip-Bu Tan says the chipmaker will expand beyond CPUs, with early GPU efforts focused on data-center workloads.
Topics
News
- Claude Owner Anthropic Discussing In-House AI Chips, Report Says
- Intel, Google Expand AI Infra Partnership with Focus on CPUs
- Anthropic Begins Project Glasswing to Hunt Software Flaws With AI
- AWS CEO Remains Bullish on Middle East Despite Drone Strikes on Data Centers
- How Google’s “Near-Accurate” AI Overviews Fuels Misinformation
- Automation Foundation Key to AI Success, Says Report
Intel said it plans to start producing graphics processing units (GPUs), marking a strategic expansion beyond its long-standing focus on central processing chips as it seeks a stronger foothold in artificial intelligence hardware.
Chief executive Lip-Bu Tan announced the effort on Tuesday at the Cisco AI Summit. He said Intel will start making GPUs, which are widely used in gaming and in training and running AI models.
“Intel will start producing GPUs,” Tan said, adding that the company will shape its approach based on customer needs.
The project is being led by Kevork Kechichian, executive vice president and head of Intel’s data center group, who joined the company last year as part of a broader effort to bolster engineering leadership.
Intel also hired Eric Demers in January after more than 13 years at Qualcomm, where he most recently served as a Senior Vice-President of Engineering.
The move marks a shift for Intel, which has long centered its business on CPUs. GPUs have become increasingly important as demand grows for computing power needed in AI systems, an area where Nvidia currently holds a strong market lead.
Tan, who became Intel’s CEO last March, said the GPU initiative will initially focus on data center workloads and that the company is already working with customers to define product requirements.
He reiterated that GPU production and strategy would evolve in response to client feedback.
