Morgan Stanley: AI Focus Shifts From Computing Power to Memory Efficiency, Emerging Memory Market to Hit $23B by 2030

According to Morgan Stanley's report released Thursday (July 16), artificial intelligence development's next critical bottleneck has shifted from computing power to memory efficiency, marking the end of the industry's pure processing capacity race. The report highlights that while memory bandwidth grows only 14% annually, AI token expansion exceeds 320-fold, creating a significant bandwidth gap.

Memory costs are rising sharply, with memory expenditures expected to account for 40% of cloud service providers' capital spending by 2027, up from 12% in 2023. Morgan Stanley estimates the emerging memory technology market (excluding HBM) will reach $23 billion by 2030, compared to $1.2 billion in 2025, signaling a major shift toward data storage and retrieval architecture innovation.