💾 AI infrastructure bottlenecks

• AI server supply chains are getting tighter, not looser. The key read-through is that shortages have spread beyond GPUs into CPUs, memory, storage, high-end PCBs, and power management, to the point that upstream GPU/CPU vendors now verify all parts are in place before shipping to server makers “to avoid wasting bullets.” That signals allocation discipline and continued capex pressure across the stack. 1

• SK Hynix cutting HBM package thermal resistance by 30% looks like a real edge, not just an engineering footnote. Better in-package cooling directly supports higher-density HBM5 stacking, and with $NVDA Rubin already tied to next-gen HBM supply, memory is increasingly part of the AI performance moat rather than a commodity sidecar. 2

• AI-driven power-spec upgrades are likely to deepen the global MLCC squeeze. Lead times already stretching to 20+ weeks suggest passive components are becoming another hidden choke point in AI hardware ramps. 3

🏭 Semiconductor supply chain pricing and expansion

• GlobalWafers saying 12-inch wafer demand has fully recovered because of AI, while discussing 2nd half price hikes, points to upstream silicon pricing power returning. The setup matters because it suggests AI demand is now strong enough to absorb cost pass-throughs in foundational semiconductor inputs. 4

• Wiwynn planning to build at least 3 more plants in the United States after demand at its new US facilities came in “beyond anything we imagined” reinforces that AI server manufacturing is being pulled onshore by customer urgency, not just geopolitics. That usually means hyperscaler demand visibility is strong enough to justify capacity adds before the cycle cools. 5

🚀 Space and defense upside

• $MDA is framed as a top long in space tech with a path to $100+ over time, driven by a defense pivot and pipeline expansion. The core angle is that the story is no longer just legacy space exposure; the upside comes from moving into higher-priority defense demand and broadening the revenue backlog. 6

📚 Trading and options education

• Learning Crypto $btc volatility trading is pitched as broadly useful for buy-side options thinking because options knowledge is interconnected rather than siloed. The practical takeaway is that vol trading sharpens understanding across structures and directional setups, instead of being a niche skill set. 7