With its AI strategy, Intel is betting on doing everything at once: rescuing x86—the dominant architecture in servers and PCs—promoting GPUs, and designing ASICs, i.e., custom chips for specific tasks. The goal is to reposition itself as a comprehensive provider.
Intel is gearing its strategy toward covering generative, inference, and physical workloads. Generative AI creates content; inference executes models in production; physical AI connects algorithms with industrial processes. This approach defines new investment and design priorities.
Intel’s AI Strategy
The company‘s revenue in 2025 totaled $52.853 billion, down 0.5% year-on-year. Operationally, lower revenue in client computing was offset by higher demand in the data center and artificial intelligence segment.
At the same time, net income reached $26 million, reversing the $19.233 billion loss in 2024. This positive result was driven by the extraordinary gain of $5.6 billion generated after the divestiture of 51% of Altera.
The company adopts a heterogeneous architecture that combines x86 CPUs with specialized accelerators. Within this framework, it launched Core Ultra for PCs with integrated AI and Xeon 6 for data centers, as well as GPUs such as Crescent Island, focused on inference efficiency.
Intel Foundry is establishing itself as a strategic pillar. It operates as a manufacturing business for third parties and for Intel itself, using advanced nodes such as Intel 18A and Foveros packaging, a key technology for integrating multiple chips into a single system.
Intel has also formed an alliance with NVIDIA to develop customized solutions that combine x86 and acceleration. At the same time, it promotes open software, optimizing frameworks such as PyTorch, widely used in AI, for its xPU portfolio.
Finally, the company prioritizes agent AI—autonomous systems that make decisions—and physical AI, geared toward real-world environments. To this end, it has unified the engineering and design of custom ASICs, seeking differentiated solutions with high added value.
Competition
In DCAI, the competition includes AMD, which rivals CPUs, GPUs, and accelerators with its x86 architecture. NVIDIA and its GPU systems also stand out, with demand growing as AI has become a central driver of computing.
The environment is expanding with hyperscalers such as Amazon, Google, Meta, and Microsoft, which develop their own silicon. In addition, players with ARM and RISC-V, and Broadcom in custom ASICs, are advancing in a landscape that will continue to intensify.