AI hardware evolution: CPUs and NPUs gain importance alongside GPUs

The role of processors in the artificial intelligence era is evolving beyond the dominance of Graphics Processing Units (GPUs). As AI shifts from simple generative responses to autonomous AI agents capable of planning, tool use, and data management, the importance of the Central Processing Unit (CPU) is increasing.

Arm suggests that AI agents will require CPUs to handle complex tasks such as code execution, service orchestration, and memory management. Similarly, NVIDIA is developing the ‘NVIDIA Vera’ CPU, specifically designed for AI agents to accelerate processing and reduce GPU idle time. This shift emphasizes a division of labor where GPUs handle massive parallel computations while CPUs manage the surrounding logic and system integration.

On the consumer side, the market for AI-capable hardware is expanding with the arrival of budget-friendly laptops. New machines from manufacturers like Asus, Acer, HP, and Lenovo are incorporating Neural Processing Units (NPUs) to run AI tasks locally. These devices, often priced under $1,000, allow for offline AI-assisted tasks such as photo editing and noise reduction, reducing reliance on cloud computing.