Samsung is reportedly working on an AI accelerator chip for PCs as part of its effort to expand dedicated on-device AI processing. Multiple reports say the chip, codenamed “Gaia,” is designed to support PC “edge AI” workloads, where AI runs locally rather than relying solely on the cloud. The chip is described as using a 4nm manufacturing process and featuring an optimized NPU (neural processing unit) architecture intended to improve efficiency for AI tasks on PCs.
The reports also indicate that Samsung has already provided prototypes to some unnamed “major clients.” In addition, Samsung is said to be planning to begin mass production of the chip next year, suggesting it is moving from early development toward commercial deployment.
Beyond PCs, the described use cases extend to other “physical AI” applications such as robotics, reflecting a broader strategy to supply AI compute hardware across multiple device categories. The news positions Samsung in a competitive field that already includes companies such as Nvidia and Qualcomm in the PC and AI accelerator market, though the reports focus primarily on Samsung’s own roadmap for Gaia.
