Jim Cramer Says Buy 2 AI Stocks up 460% and 1,300% Since 2023 -- Wall Street Agrees

Street upside remains compelling for AI leaders: Nvidia’s median target is $300 vs. $211 current (~42% upside) and Meta’s is $815 vs. $669 (~22% upside), with both framed as benefiting from AI demand. The article highlights Nvidia’s >80% AI accelerator market share and full-stack platform, while Meta plans to monetize AI infrastructure via a cloud business with AI-model/API and capacity offerings, after $111B capex in 2024-25 and guidance toward $135B in 2026.

Analysis

The main market error here is treating the AI capex cycle as a binary winner-take-all narrative. For NVDA, the real variable is not whether hyperscalers build some custom silicon; it is whether frontier workloads still require enough NVIDIA-standardized systems to keep cluster spend compounding. That argues for persistence in revenue momentum over the next 2-3 quarters, but the stock is now more sensitive to any sign of order digestion or mix shift than to headline competition.

META is a different mechanism entirely: if it can monetize spare compute, investors may start underwriting its data center buildout as an asset business rather than a pure expense line. The second-order risk is that the market eventually realizes external cloud sales can be low-margin, operationally messy, and potentially distract from the high-ROIC ads franchise; in that case, the multiple expansion case weakens even if revenue grows. The bullish setup is strongest only if Meta proves utilization with minimal incremental capex and no deterioration in ad margins.

The contrarian view is that the consensus is probably overreading the cloud angle and underpricing execution friction. A nascent cloud business is not the same as a durable platform, and the stock reaction could be ahead of the financial disclosure. The cleaner trade is still relative: NVDA remains the higher-quality exposure to continued AI spend, while META is more of a self-help story that needs evidence, not commentary. Falsifiers: a capex step-down or margin compression at the next earnings prints, or any guidance implying AI infrastructure payback is lengthening rather than shortening.

Key Decisions for Investors

• Stay long NVDA on any 3-5% pullback into earnings or guidance events; use a 1-3 month horizon and exit if management signals order normalization, gross margin pressure, or a slower-than-expected AI server ramp.
• Do not chase META on the cloud headline alone; only add on confirmation that third-party compute revenue is material and that ad margins remain intact. Otherwise treat it as a watch item, not a conviction long.
• Relative-value idea: long NVDA / short GOOGL or MSFT only if upcoming cloud commentary shows capex returns deteriorating or hyperscaler AI spending growth slowing; this pair is a 3-6 month trade, not a day trade.
• If META rallies on the cloud story before there is operating disclosure, consider selling upside via call spreads rather than owning outright; the story is plausible but still under-validated.
• Watch next earnings for a split between capital intensity and monetization. If META capex keeps rising without clear external revenue and NVDA order visibility stays tight, the better risk/reward remains NVDA over META.