Intel has officially announced plans to begin producing its own graphics processing units (GPUs) — a major strategic move into a market long dominated by Nvidia’s chips used in gaming, artificial intelligence, and data‑center applications.
At the Cisco AI Summit on February 3, 2026, Intel CEO Lip‑Bu Tan confirmed that the company will develop GPUs as part of its expanding semiconductor strategy, moving beyond its traditional focus on central processing units (CPUs). As Tan explained, Intel will align its GPU development around customer needs, particularly for data‑center AI workloads where Nvidia currently holds substantial market share.
New Leadership and Strategic Focus
To lead this effort, Intel has hired Eric Demmers, a former Qualcomm executive with extensive GPU engineering experience. Demmers will work under Intel’s data‑center chip division, overseen by Kevork Kechichian, signaling the company’s intent to target enterprise and AI applications rather than solely consumer graphics.
The announcement suggests that Intel is positioning its GPU business not just as an alternative for gaming but as a competitor in the AI and data‑center accelerator space — historically dominated by Nvidia’s architectures. While Nvidia didn’t invent the GPU, its products have become essential for large‑scale model training and inference applications.
Early Stages with Customer Engagement
Intel describes the initiative as early stage, emphasizing that its GPU architecture and production timelines will be shaped in part by feedback from early customers. Tan noted that some clients are already actively engaging with Intel’s 14A manufacturing technology, which could become a foundation for volume GPU production later in 2026.
Industry observers see this move as part of a broader chip industry effort to diversify supply chains, counter Nvidia’s commanding position, and tap into the rapidly growing AI hardware market.
0 Comments