"Internet-famous Chip Analyst" P Equity Interview: Compute Shortage, Memory Cycle, Hidden Bottlenecks, Copper Out / Optics In, and the "Big Lie of 10-Year Demand Visibility"
Summary
The Wall Street Journal Chinese edition (华尔街见闻) published an interview with the "internet-famous chip analyst" P Equity, covering the compute supply-demand gap, the memory cycle, optical replacing copper interconnects and other semiconductor supply-chain bottlenecks, while calling the industry's claimed "10-year demand visibility" a big lie. The interview offers no concrete capex, funding or valuation figures; it is a skeptical commentary on the AI compute demand narrative itself.
Bubble analysis
This interview matters to the bubble question because it directly challenges the core assumption underpinning AI investment: demand visibility. If companies are planning enormous capex on the premise that compute demand is assured for a decade, and an industry analyst calls that premise a lie, then today's spending may rest on over-optimistic demand forecasts — a key mechanism in the bubble argument. But the piece contains no figures, so it is only indirect evidence.
Related signals
linked by shared tags
- CICC: Research Methods and Insights on AI Compute Hardware — Compute Chips, Storage, Equipment and Consumer Electronics, Sept 2026#compute#chips
- Why Is DeepSeek Embracing Huawei Now?#compute#chips
- Micron says robots will keep RAM prices high, even if the AI bubble pops#chips#memory
- South Korea's September Chip Exports Surge 263% Year-on-Year to a Record $60.3 Billion#chips#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#chips#memory