Robots Emerge as Next Growth Driver for Samsung, SK hynix

The projected increase in memory demand stems from two main factors. First, humanoid robots are beginning to integrate 'robot foundation models,' high-performance AI brains that require greater DRAM capacity per unit. Counterpoint Research anticipates the DRAM capacity per humanoid robot will more than double, from 19 gigabytes this year to 39 gigabytes by 2030.

Second, global humanoid robot shipments are forecast to grow substantially, from 60,000 units this year to approximately 500,000 units by 2030, as reported by Smart Analytics Global. This combined growth in per-unit capacity and overall shipments translates into a significant increase in total memory consumption for the sector, rising from roughly 8,300 terabits this year to 172,000 terabits in 2030.

Memory manufacturers are preparing for this emerging demand. Samsung Electronics stated in July it was "preemptively preparing next-generation robot semiconductor solutions" including high-bandwidth low-power memory and advanced packaging. SK hynix plans to jointly develop next-generation memory for Nvidia's Jetson Thor robot chip platform. Industry officials note this shift will particularly boost demand for low-power DRAM (LPDDR), critical for robot performance, and high-capacity NAND flash.