Jim Cramer Says Buy 2 Artificial Intelligence (AI) Stocks Up 875% and 1,400% Since 2023

Nvidia reported fiscal Q2 2027 revenue growth of 106% to $96.2B and non-GAAP EPS growth of 120% to $2.22, while Wall Street projects 72% annual adjusted-earnings growth through fiscal 2028. The article argues Nvidia's 36x forward P/E and 0.5 PEG support a buy case, with a $318 median analyst target implying 42% upside from $223. CrowdStrike grew fiscal Q2 revenue 26% to $1.4B and non-GAAP EPS 35% to $0.31, but its 230x earnings multiple is viewed as excessive after a 115% YTD rally; its $245 median target implies 4% downside from $255.

Analysis

NVDA’s valuation debate is no longer about near-term demand visibility; it is about whether hyperscaler AI capex converts from a concentrated training buildout into recurring inference and enterprise deployment. The second-order upside is in networking, systems and software attach, where a larger share of customer spend can offset lower accelerator pricing over the next 6-18 months. The key falsifier is not a modest GPU shipment miss, but a sequential reduction in cloud-provider capex plans or gross-margin pressure that signals competitive/custom-silicon substitution is reaching economic scale.

CRWD is more exposed to multiple compression than to a deterioration in its operating franchise. At its current earnings multiple, even sustained high-20% earnings growth leaves little room for billings deceleration, weaker net-new module adoption, or elevated sales-and-marketing spend; a one-turn reset in long-term growth expectations can produce an outsized drawdown within 1-3 months. Microsoft and PANW remain the relevant competitive risks because broader platform bundling can force higher discounting or slow endpoint-to-cloud/identity cross-sell, even if endpoint retention stays strong.

The contrarian view is that the market may be treating both names as interchangeable AI beneficiaries. NVDA is primarily a capex-cycle and supply-chain-duration exposure, while CRWD is a high-duration software multiple whose AI benefit is harder to monetize independently of normal security consolidation. That supports owning NVDA selectively on volatility while using CRWD as a relative-value funding leg rather than making a broad bearish call on cybersecurity demand.

Key Decisions for Investors

• Initiate or add to NVDA on an 8-12% market-driven pullback; target a 15-25% return over 6-12 months if hyperscaler capex guidance remains intact. Reduce if the next two major cloud earnings cycles show aggregate AI-capex cuts or if NVDA guides to meaningful sequential gross-margin erosion.
• Establish a 3-6 month pair: long NVDA / short CRWD, sized beta-neutral. The thesis is multiple dispersion rather than absolute AI direction; target 10-15% relative performance, with a stop if CRWD reaccelerates billings materially above revenue growth while NVDA’s data-center growth decelerates sharply.
• For cybersecurity exposure, prefer PANW over CRWD for the next 6-12 months where valuation discipline is required; PANW’s platform consolidation opportunity offers similar security-spending exposure with less dependence on a premium terminal multiple. Reassess if CRWD demonstrates sustained acceleration in identity/cloud module adoption and operating-margin expansion.
• Set an event alert around quarterly disclosures from MSFT, AMZN, GOOGL and META: capex guidance is the highest-frequency validation point for NVDA, while security-platform pricing commentary and large-deal duration are the cleaner leading indicators for CRWD’s downside risk.