How is the role of AI PCs changing?

  • Gartner expects global AI PC shipments to reach 143 million units in 2026, accounting for 55% of the entire PC market, with shipments exceeding those of traditional PCs.
  • Traditional PCs primarily provide computing services, while AI PCs place greater emphasis on enabling personal agents to understand natural-language instructions and help complete tasks. End-user devices store a substantial amount of personal information; local execution can help deliver personalized services and protect privacy.

What hardware needs will local model execution change?

  • Following NVIDIA’s release of RTX Spark, Jinyuan Securities believes AI agents are gradually moving from cloud services to local devices, and that PCs will evolve from office computers equipped with neural processing units into platforms capable of running large language models and agents locally.
  • According to the firm’s analysis, demand for local computing is driving configuration upgrades. High-computing-power chips, large-capacity unified memory, advanced packaging, thermal management, and power-supply management are among the areas that could benefit. Coordination between local and cloud models could also improve response speed, privacy and security, and usage costs.

What business connections do the related companies currently have?

  • ThunderSoft provides full-stack AI PC products and technology, maintains cooperation with supply-chain partners and leading PC manufacturers, and has previously launched several AI PC products.
  • Everwin Precision’s consumer electronics business revenue in the first half was 5.758 billion yuan, up 8.59% year over year, with AI PC-related revenue increasing markedly. Adoption in areas such as personal office work, content creation, software development and programming, and cross-application automation will affect demand for device upgrades.