Memory Takes Over from Compute: Six Major Semiconductor Innovation Themes in the Second Half of the AI Era
Summary
An article in EE Times China argues that as AI compute demand keeps expanding, the semiconductor industry's investment and innovation focus is shifting from compute chips such as GPUs toward memory (HBM, DRAM and related), outlining six major semiconductor innovation themes for the second half of the AI era. It is an industry technology-trend piece and provides no specific capex, funding or valuation figures.
Bubble analysis
This article is about the evolution of semiconductor technology roadmaps (memory taking over from compute), not about whether AI investment is overheated. It offers no capex, funding or valuation figures and does not weigh AI revenue against AI spending, so it speaks only indirectly to the bubble question — at most it suggests the supply chain is still expanding capacity for AI demand, which is weak evidence.
Related signals
linked by shared tags
- Huatai Securities: AI Supply Chain Volume Drives Korea's Exports Higher#memory
- Micron says robots will keep RAM prices high, even if the AI bubble pops#memory
- South Korea's September Chip Exports Surge 263% Year-on-Year to a Record $60.3 Billion#memory
- Don't Expect Cheaper Memory Even If The AI Bubble Pops: Micron Says Each Humanoid Robot Would Need Over 200 GB Of DRAM And Multiple Terabytes Of NAND#memory
- Micron FY2026 Annual Report: Net Income Attributable to Parent Reaches $84.969 Billion, Up 895.07% Year-on-Year#memory