Generative AI could eventually transition to physical artificial intelligence, resulting in the mass production of humanoids that can make decisions based on proactive behavior. This is expected to become possible thanks to Micron’s Edge Inference Memory (EIM), which aims to deliver the same low-power properties as LPDDR while delivering the low-latency and high-bandwidth attributes of HBM. Best of all, EIM could arrive sooner rather than later, with a report stating that commercialization may happen as soon as 2028. EIM could also be adopted in LPDDR6 memory, and without the increased power draw of stacked HBM AI-powered robots will have significantly […]
Read full article at https://wccftech.com/micron-edge-inference-memory-hbm-power-efficiency-ai-robots/