Unified memory lifts AI PC ambitions as Chinese open models gain local-inference ground

PC vendors are expanding memory capacity while software platforms make it easier to run larger open-weight models directly on devices. The boundary between cloud-scale AI and local inference is blurring, helped by Chinese open models.

• PC vendors are increasing memory capacity for running AI models locally
• Software platforms simplify running large open-weight models on devices
• Chinese open models strengthen their position in local inference
• The line between cloud and local AI is becoming increasingly porous