Micron Developing EIM Memory for Humanoid Robots

What happened

Lee Cheol, vice president at Micron, said at the "Semiconductors Powering Humanoids" conference held as part of SEDEX 2026 at COEX in Seoul on Oct. 16 that Micron is preparing edge inference memory (EIM) for physical AI applications and is discussing with major customers the development of products suited to emerging markets including autonomous driving, humanoid robots, AR and VR. EIM is positioned between AI chips and memory to reduce data processing latency and increase bandwidth; it stacks LPDDR on top of a logic die and, unlike PIM, its computing architecture has not been disclosed, and Micron has not disclosed specific performance targets. The company is developing a method of applying EIM to LPDDR6, with industry observers expecting mass production in the second half of 2028. Micron noted that current AI models for robots use around 1 billion to 2 billion parameters, so memory bandwidth requirements are not particularly high, but as more advanced vision-language models (VLMs) emerge, LPDDR alone will become insufficient, requiring a transition to HBM, which Micron expects to eventually be adopted in humanoid robots. Micron expects future high-performance humanoid robots to require approximately 200GB of DRAM and 4TB of storage, compared with 12GB of DRAM and 256GB of storage used in current service robots. The company said it is collaborating with robot manufacturers, chipset companies and hyperscalers from the early design stages.

Impact and signal

Affected components and suppliers

Source and time

Source: The Elec · 10/17/2026 · View original source