2 Artificial Intelligence (AI) Stocks That Can Skyrocket Up to 484%, According to Select Wall Street Analysts

Analysts tout major upside for AI-linked names: AMD is projected to gain up to 164% based on a Baird price target raised to $1,250 (from $625), underpinned by potential share gains in AI data center GPUs. Separately, SpaceX (SPCX) is modeled to rise as much as 484% by 2031 to $800 per share, with projections of $837B sales in 2031 and ~$696B EBITDA. Despite the bullish targets, the article flags valuation skepticism and funding/delivery risks for SpaceX’s expansion.

Analysis

The actionable read-through is not the eye-popping targets; it is that capital is still willing to pay for a credible second source in AI compute. AMD’s upside depends on proving it can win sockets without turning that win into margin leakage; if hyperscalers use AMD mainly as pricing leverage against NVDA, revenue can move faster than earnings and the stock can still disappoint. The cleaner beneficiaries of any broadening AI capex remain the picks-and-shovels layer—advanced packaging, foundry, and networking—because share shifts still require wafers, HBM, and rack buildouts regardless of which GPU wins.

For NVDA, this is a medium-term multiple risk more than a near-term demand risk. The first falsifier is next-quarter hyperscaler capex commentary and AMD’s own datacenter mix: if those do not confirm broader diversification, the market will fade the narrative within 1-3 months. Over 6-18 months, the bigger risk is industry ASP compression as customers force a two-horse race; that helps end users, but it can cap the incremental earnings power of both GPU vendors.

SpaceX is not a clean public-market expression, which is the point: the valuation case assumes years of uninterrupted financing and execution, so the equity story is really a leverage-and-dilution story in disguise. If funding needs rise faster than operating cash flow, the economic winners are lenders and select suppliers, not common equity. The contrarian miss here is that ‘AI TAM’ often gets priced before returns on capital are visible, while funding risk and execution slippage get pushed into later rounds.

Key Decisions for Investors

• Long AMD / short NVDA as a 1-3 month relative-value trade only on pullbacks, targeting 10-15% relative outperformance if AMD’s next datacenter guide confirms acceleration; stop if AMD gross margin or AI backlog disappoints.
• Prefer SOXX or SMH over single-name chasing on any AI pullback; this captures second-order winners such as TSM, AVGO, ANET, and memory suppliers with less execution risk than an outright AMD position.
• Do not use TSLA as a proxy for SpaceX upside. Treat the SpaceX valuation story as non-actionable for public equity until there is a financing, listing, or major operating milestone that changes dilution risk.
• If you want bearish exposure, sell AMD upside via call spreads after a sentiment spike rather than shorting outright; the setup is execution-sensitive, and the market may reward momentum before fundamentals catch up.