The Best Artificial Intelligence (AI) Growth Stocks on the Nasdaq That Wall Street Loves Right Now

The article is broadly bullish on two AI stocks: Applied Digital and Nvidia. Applied Digital is expected to build out four new AI data centers and target $1 billion in NOI over five years, while 100% of 13 analysts rate it a buy with a $43 median target, implying 32% upside. Nvidia is described as a value at 24x forward P/E and 0.72 five-year PEG, with 93% of 70 analysts rating it a buy and a $267.50 median target suggesting about 33% upside.

Analysis

The market is quietly separating AI infrastructure from AI compute: the former is becoming the bottleneck trade, while the latter is already a consensus winner. That creates a second-order setup where capacity owners with signed anchor tenants can re-rate faster than the hardware complex, but only if execution risk stays contained through build-out and financing windows. APLD sits in that sweet spot, but the gap between contracted demand and actual cash flow is still wide enough that any slippage in energization or tenant conversion could reset expectations hard.

NVDA is the cleaner expression of the same theme, but the more interesting point is that its valuation reset may actually extend the duration of the cycle rather than end it. If the stock is now being underwritten on a market-multiple forward P/E instead of a scarcity premium, funds that previously avoided the name for valuation reasons may re-enter mechanically, supporting flows for months. The key risk is not demand softness; it is whether hyperscalers keep spend disciplined enough to avoid a digestion phase that compresses server and networking orders after the next wave of capacity comes online.

The consensus is underestimating how much beneficiary dispersion there is beneath these two names. Oracle and CoreWeave are indirect winners from data-center availability, but they also become bargaining anchors for future capacity pricing, which could pressure margins for any operator that needs to refinance or renegotiate power terms later. The contrarian read: the market may be overpaying for the right to own the obvious AI winners and underpricing the infrastructure peers with less headline appeal but better cash conversion and shorter catalyst windows.

Near term, APLD is a months-to-years story with binary financing/execution checkpoints, while NVDA is a days-to-months sentiment and flow trade layered on top of years-long earnings power. The main reversal triggers are macro risk-off, capex pause announcements from hyperscalers, or a build-out miss that exposes how much of APLD’s upside is still option value. If those do not materialize, the likely path is continued multiple support for NVDA and an increasingly crowded but still workable re-rating in select infrastructure names.

Key Decisions for Investors

• Long NVDA on pullbacks over the next 2-6 weeks; target a low-to-mid teens upside move if forward multiples remain anchored near market and hyperscaler capex stays intact. Risk: de-rating if AI spend normalizes or broader tech multiples compress.
• Initiate a staged long APLD position only on weakness, not strength, and size it as a venture-style exposure for a 6-18 month horizon. Best risk/reward is if the market continues to pay for contracted capacity optionality before NOI shows up in earnings.
• Pair trade: long NVDA / short a basket of overextended AI hardware names with weaker balance sheets over 1-3 months. This isolates the best-in-class earnings machine against names where valuation is still doing too much of the work.
• Consider a small long ORCL / short APLD relative-value trade if you want AI infrastructure exposure with lower execution risk. ORCL has less upside torque, but the path to monetization is cleaner and should defend better in a risk-off tape.
• Buy NVDA call spreads rather than outright stock if you expect continued upside but want to reduce premium paid after the recent rerating. Best use case is a 1-3 month catalyst window with asymmetric upside from renewed flow support.