3 Stocks to Buy Following SpaceX's First Earnings Call

Nvidia is set to be SpaceX’s exclusive chip provider, including for the Vera Rubin Starmind AI satellite compute payload—an incremental but concrete demand signal via Nvidia GPUs/CPUs, with prototype delivery early next year and mass production targeted for 2027. Musk’s call argued DRAM output rises ~20%/yr while demand rises ~200%/yr, implying sustained supply tightness and higher memory prices—supportive for SK Hynix and Micron. The article cites SpaceX AI capex of $15.8B in Q2 ($18.4B total) and notes SK Hynix’s >50% HBM market share plus a reported $500B multiyear supply/co-development deal for Vera Rubin memory.

Analysis

The market implication is less about a single design win and more about the validation of AI compute demand outside the hyperscaler cohort. That broadens the addressable narrative for NVDA, but the cash-flow impact is delayed enough that the immediate effect is mostly multiple support; the tradeable catalyst is sentiment, not revenue.

Memory looks cleaner than compute here. If AI workloads keep pushing HBM/DRAM intensity faster than wafer supply, the operating leverage sits with SKHY first and MU second because pricing power accrues before volumes do. A secondary effect is that higher memory content raises bill-of-materials costs for accelerator and server OEMs, which can either delay purchases or force product mix downshifts if customers become more price sensitive.

Contrarian view: the consensus is likely overestimating the near-term monetization of this announcement and underestimating how much of the upside is already in NVDA’s consensus. The key falsifier is not the headline, but whether spot DRAM/HBM pricing and Nvidia’s forward commentary stay tight over the next 1-2 quarters; if lead times shorten or memory capex normalizes, the memory supercycle thesis loses urgency quickly. The 6-18 month structural risk is that Nvidia’s roadmap reduces memory intensity per unit of compute, which would cap the upside for the suppliers even if AI demand remains strong.

Key Decisions for Investors

• Long SKHY vs. NVDA on a relative-value basis for 3-6 months: the thesis is that HBM scarcity monetizes faster than NVDA’s already-rich AI narrative. Stop if NVDA re-accelerates guidance materially or if SKHY loses pricing power in channel checks.
• Add to MU on pullbacks over the next 1-2 weeks, targeting the next DRAM contract-pricing leg higher. Falsifier: any sign that consumer/enterprise DRAM inventory is rebuilding or that spot prices roll over by high-single digits.
• Own NVDA only on weakness rather than chasing the headline, because the incremental earnings contribution from this project is likely FY27/FY28+ optionality. Risk/reward is better after a 3-5% pullback or on a broader semis de-risking day.
• Watch item: if HBM lead times and spot DRAM keep tightening for another quarter, extend SKHY/MU exposure; if they stabilize, take profits quickly because the scarcity premium can unwind faster than the equity multiple.