fizzzi
06-08

$Roundhill Memory ETF(DRAM)$ $SanDisk Corp.(SNDK)$ $Taiwan Semiconductor Manufacturing(TSM)$ Elon just made it crystal clear: new fabs, even the best ones planned in the US, won’t come close to covering memory demand over the next 5 years. The Idaho and New York fabs? Earliest volume is 2028–2030, just a tiny fraction of what AI compute needs.

Demand for AI logic, memory, and packaging is exploding, and supply simply won’t keep up. That’s exactly why the Terafab project exists. Without it, the AI chip shortage is going to get tough. Stocks like $Micron Technology(MU)$  are looking at a potential trillion-dollar-plus market narrative if this plays out.

This isn’t hype. It’s structural.

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Comments

  • 北极篂
    06-09
    北极篂
    不过，我认为市场最大风险是「预期跑太前面」。半导体终究是周期行业，如果未来两三年供给快速释放，估值也可能重新修正。但至少现阶段，我更倾向把这看成结构性趋势，而不只是短期炒作。因为AI真正缺的，已经不只是算力，而是整个供应链产能。
  • 北极篂
    06-09
    北极篂
    我个人觉得，市场可能还是低估了AI对内存和先进封装的需求爆发速度。过去大家都把焦点放在GPU，但其实真正限制AI扩张的，越来越像是HBM、DRAM和先进封装产能。如果未来5年晶圆厂扩产速度真的追不上需求，那受益的不只是芯片设计公司，而是整条供应链，尤其是内存和代工。像美光，最近股价涨很多，但如果HBM持续供不应求，现在的估值未必算贵；台积电的先进封装能力，也可能成为AI时代最稀缺的资源之一。
Leave a comment
2