The Nasdaq Is on Fire. Here Are the 2 Best Artificial Intelligence (AI) Growth Stocks That Still Look Cheap.

The article argues that Nvidia remains relatively cheap at 41x earnings despite fiscal 2026 revenue of $216 billion (+65% YoY) and net income of $120 billion (+65%), while its $4.9 trillion market cap may cap near-term multiple expansion. CoreWeave is highlighted as a higher-risk AI infrastructure name with $5.1 billion in 2025 revenue (+168% YoY), a $66.8 billion backlog, and a 10x sales multiple, but it carries $21.4 billion of debt against $3.3 billion in book value and posted a $1.22 billion loss. Overall tone is constructive on AI fundamentals but cautious on valuation and leverage.

Analysis

The market is implicitly treating NVDA as a mature mega-cap rather than a hyper-growth platform, which is why the multiple compresses despite exceptional operating momentum. The second-order effect is that the marginal AI dollar is increasingly migrating from semiconductor beta into infrastructure beta: if compute demand keeps compounding, the true bottleneck shifts from chips to deployment capacity, networking, and power, which should continue to support the broader AI supply chain even if NVDA’s upside becomes more “earnings compounding” than “multiple expansion.”

CRWV is a classic balance-sheet call option on AI capacity demand. The enterprise value is being priced as if backlog converts cleanly into margin, but the debt load means equity holders are effectively long execution and refinancing conditions at the same time; that makes the stock highly sensitive to any slippage in utilization, receivables collection, or capex intensity over the next 2-4 quarters. In contrast to NVDA, the upside is less about steady growth and more about convexity if the market starts believing debt-funded scale can translate into free cash flow.

The contrarian miss is that “cheap” in AI may not mean undervalued so much as duration-mismatched. NVDA’s size limits multiple re-rating, but its earnings durability could make it a lower-volatility AI core holding; CRWV may look optically inexpensive until the market demands proof of financing resilience. A broad AI pullback would likely punish CRWV first, while NVDA would probably act as the defensive beneficiary as capital rotates to the highest-quality AI compounder.

Catalyst-wise, the next 1-2 quarters are all about guidance quality, backlog conversion, and incremental debt markets. If CRWV can show accelerating gross margin and lower funding spread, the equity could re-rate sharply; if not, the leverage structure becomes the dominant narrative. NVDA’s key risk is not near-term demand, but a slower-than-expected pace of incremental upside versus consensus, which could cap further multiple expansion even with strong fundamentals.

Key Decisions for Investors

• Maintain a core long NVDA vs. short basket of lower-quality AI infrastructure names over the next 3-6 months; NVDA should outperform on any sector de-risking because its earnings visibility is materially better.
• For higher-risk capital, consider a tactical long CRWV only on post-earnings confirmation of margin improvement and debt-market stability; use a 1-2 quarter horizon and size as a volatility trade, not a fundamental compounder.
• Pair trade: long NVDA / short CRWV to express ‘quality within AI’ dispersion; target a 15-25% relative outperformance if the market re-prices leverage risk and rewards cash-generation durability.
• Avoid adding to CRWV ahead of refinancing or capex-heavy updates; the stock is most vulnerable if the market shifts from backlog excitement to balance-sheet scrutiny.
• If NVDA trades down on any broad AI rotation without a change in fundamentals, use weakness to add via staged entries over 2-4 weeks; downside should be better supported than the cohort because the market treats it as the cleanest AI liquidity sink.