Prediction: 1 Artificial Intelligence (AI) Stock That Will Be Worth More Than Palantir and Micron Combined by 2028

AMD could see major AI-driven revenue upside from two 6 GW GPU deals with OpenAI and Meta, with the article estimating roughly $80 billion of long-run revenue potential from each deal. Analysts expect AMD revenue to rise 35% in 2025 to $46.6 billion, with earnings growth accelerating as AI data center deployments begin in 2H 2026. The piece is comparatively bearish on Micron and Palantir due to valuation and demand-risk concerns, but the overall message is constructive for AMD.

Analysis

The market is treating this as a simple ‘AI winners vs losers’ story, but the more important dynamic is capital intensity migration. If hyperscalers and model builders keep pouring incremental capex into GPUs and custom silicon, the marginal beneficiaries shift upstream from software narratives like PLTR toward hardware with real dispatchable revenue visibility. AMD’s edge is not just product share gain; it is that each design win converts into multi-year supply commitments, which creates a rarer earnings path in semis: visible top-line compounding plus multiple support from scarcity of credible alternatives.

The main second-order effect is pressure on the AI infrastructure stack. A sustained AMD ramp would squeeze gross profit pools at incumbent GPU suppliers and force a response in pricing, packaging, and software incentives. That matters because the market is currently underwriting some names on the assumption that AI demand stays concentrated and uneconomic substitutes remain distant; if AMD proves it can take share in production deployments, the valuation gap versus the rest of the AI cohort should narrow quickly.

The contrarian risk is that the bull case is too linear on deployment timing. These contracts are big but execution-sensitive, and the revenue recognition profile can slip materially if power, data center buildout, or software integration delays push first deployments right by 2-4 quarters. In that case, the stock could de-rate even while the long-term thesis remains intact, because the near-term multiple is already discounting a clean acceleration path.

For MU, the risk is that the AI memory thesis gets treated as perpetual when it is really cyclical around a few platform transitions. If the market starts to believe memory content per rack is peaking, the multiple can compress before earnings actually roll over. For PLTR, the issue is less fundamental growth and more duration risk: at this valuation, any slowdown in net new contract momentum or margin expansion can cut multiple faster than fundamentals can catch up.

Key Decisions for Investors

• Long AMD into any 5-8% post-news pullback; target 12-month upside is better than 2:1 if the market re-rates it from ‘growth at any price’ to ‘visible earnings compounding’
• Buy AMD 6-12 month call spreads financed with out-of-the-money puts on PLTR for a relative-value AI basket; the pair benefits if hardware monetization keeps outperforming narrative software
• Short PLTR on strength or via put spreads into earnings/contract updates; valuation leaves little room for even modest execution misses, with asymmetric downside over the next 3-6 months
• Use MU as a tactical short against AMD only if memory commentary confirms demand deceleration; otherwise stay neutral, because MU’s setup is more cyclical than structurally broken and timing risk is high
• Overweight META versus GOOGL on AI capex monetization; META has clearer near-term product-level compute demand, while the GOOGL drawback is that a breakthrough reducing memory demand could also slow broader infrastructure spend