I think the bigger picture story is that, the hyper-growth driven by non-stop spending in AI inrastructure is kind of a rear-mirror thing now. Going forward, the winner in this space will be re-defined. NVDA need to quickly find a new market beyond just datacenter GPU to sustain its hyper-growth. I think they are still fine for now with Blackwell, but if I dont see a new market developed by them by mid 2025, I may start to cut postition significantly as they will NOT be able to sustain this growth into 2026 with datacenter GPU only
今年至今的图已经显示半导体指数快进入了熊市的区域：
2年的图显示半导体指数还没有进入熊市：
半导体整个行业的P/E相对其他科技股并不高，但是也处于历史上的高点。
投资不是非白即黑，只是一种趋势的看法。这说明很多机构已经撤出半导体行业在边上观望。
昨天，整个半导体都跌，全靠英伟达一家撑起来，顶在那里。
半导体行业还有多少空间上升空间继续涨？不知道。
为什么会这样？不知道。
我已经在9月份撤出半导体ETF。
这里不是预测，只是一种观察，绝对没有投资或不投资的建议。
YMYD。
I think the bigger picture story is that, the hyper-growth driven by non-stop spending in AI inrastructure is kind of a rear-mirror thing now. Going forward, the winner in this space will be re-defined. NVDA need to quickly find a new market beyond just datacenter GPU to sustain its hyper-growth. I think they are still fine for now with Blackwell, but if I dont see a new market developed by them by mid 2025, I may start to cut postition significantly as they will NOT be able to sustain this growth into 2026 with datacenter GPU only