If I Had a Choice to Invest $10,000 in SpaceX or Nvidia, Here's What I Would Do

SpaceX reported 92% year-over-year revenue growth and a net loss that nearly halved to $541 million; Starlink generated $1.7 billion in operating income, up 79%, and doubled subscribers to 12 million. The article favors Nvidia for a $10,000 investment, citing fiscal Q2 revenue more than doubling to $96.2 billion, gross margin rising to 75% from 72.4%, and a valuation of 30 times trailing earnings, versus SpaceX at 69 times trailing sales. It flags SpaceX’s high valuation and execution risks despite its growth opportunities.

Analysis

The more useful signal is not the headline valuation comparison but the difference in earnings visibility and who bears execution risk. NVDA monetizes AI demand through an established hardware/software stack; SPCX’s prospective AI-in-space economics add substantial buildout, launch, power, cooling and utilization hurdles before headline addressable markets become cash flow. The implied customer commitments should not be treated as independent validation of end demand: spending across chip vendors, cloud providers and AI operators can be mutually reinforcing, but also exposes the chain to a capex pause or weak returns on deployed compute. CRWV’s chip purchasing and NVDA investment illustrate that counterparty and funding feedback loop; AMZN benefits from cloud demand but also faces the question of whether AI infrastructure earns adequate returns.

Near term, NVDA can benefit if supply remains the constraint, but that same constraint makes upside dependent on conversion of backlog into shipments and customer returns. Over 1–3 months, watch NVDA guidance, gross margin, delivery capacity and hyperscaler capex commentary; margin deterioration or capex deferrals would challenge the bullish case. Over 6–18 months, accelerated supply and competing/custom silicon could reduce pricing power. SPCX’s longer-dated AI ambitions may be underwritten too early: verify reported segment economics, capex, financing, and whether claimed contracts are binding and incremental before assigning value. Trailing sales and earnings multiples are not directly comparable, so the article’s valuation contrast alone does not establish relative cheapness.

Key Decisions for Investors

• Relative-value watch: consider a modest long NVDA / short SPCX expression only after independently verifying SPCX’s financials, float/liquidity and contract terms. The thesis is nearer-term earnings visibility versus execution-heavy expectations, not a clean valuation-arbitrage; cap downside and avoid sizing as a low-risk pair.
• For NVDA, add only on evidence that shipments and gross margins hold through the next guidance cycle. Falsifiers: material margin compression, order cancellations, or hyperscaler capex cuts; these would weaken both earnings durability and the multiple.
• Treat the SPCX AI/data-center narrative as an alert, not a standalone buy catalyst, until filings clarify required investment, cash generation and binding customer commitments. A sustained revenue ramp without disproportionate capex would challenge the cautious view.
• Monitor CRWV and AMZN commentary for customer concentration, utilization and returns on AI infrastructure. Weak utilization or financing stress could transmit to NVDA demand even if near-term chip orders remain strong.