

long huang
5.4K posts





$MU $SNDK $DRAM $AMD $INTC $TSM Rise of Agentic AI - Morgan Stanley's Key Takeaways -Memory becomes a key monetization leg, with agentic AI driving 26-77% of incremental DRAM demand by 2030. -Agentic AI widens the trade beyond GPUs, with CPUs becoming the control plane for multi-step workflows and system orchestration. -We estimate $32.5-60bn incremental CPU TAM by 2030 within a total server CPU TAM of $100bn+, driven by agentic workloads. -Supply-constrained enablers (foundry, ABF, BMC, interconnect) should capture outsized economics as system complexity rises. -The beneficiaries are full-stack and global, spanning CPU, memory, substrates, infrastructure ICs and equipment.






My portfolio just hit a new all-time high today. YTD: +49.32% 1Y: +191.96% The craziest part? Just over a month ago, it was still red YTD. While the YTD return is not as high as some other accounts, I am still happy because I did not have to own speculative names that went up 10x in a few months or buy too many risky short term options. This time, $MU truly made the difference, along with a few quick small bets with earlier this year like $AAOI, $NBIS, $CRWV, $ALPD, and $IREN. Thanks to the bears for creating the opportunity with Micron. How did MU crash after the last earnings, and then crash again after the Google TurboQuant panic? I still have no idea, my position size changed so quickly because of the price drop. I made some mistakes too. Took some small losses in FIG, and didn’t make much from ORCL. Still green, but the opportunity cost was high. The return could have been even higher if I hadn’t bought too much too early before the dip. But I did buy a lot near $350. It also could have been higher if I hadn’t chickened out a bit. I didn’t expect TACO to come so quickly. It could have been much higher if I had used more margin or gone heavier into call options. I only bought some calls for fun. So far this year, the portfolio is still running at roughly 3.7x QQQ’s performance. I shared my thesis with paid members during the crash, but honestly, it wasn’t as easy as it looks now.

Memory stocks experienced a DeepSeek moment. I posted about it quite early, but the market preferred to react to a few seconds of fear-driven headlines. The market was also worried about a drop in spot prices, but note that spot pricing is still very high compared to contract pricing. Read more on my portfolio updates - memory stock crash and dram demand models: theaiinvestorx.substack.com/p/portfolio-up…
