2 Artificial Intelligence Stocks That Can Have Their Nvidia Moment in 2026

CoreWeave reported Q3 2025 revenue of nearly $1.4 billion, up 134% year-over-year, with net losses narrowing to $110 million from $389 million a year earlier, about $1.9 billion in liquidity and a trailing P/S around 7—signaling rapid AI-driven top-line growth but continued cash burn. AMD posted Q3 2025 revenue of over $9.2 billion, up 36% with data center representing ~47% of sales and net income exceeding $1.2 billion (up 61% YoY); management projects a 35% revenue CAGR over the next 3–5 years, expects >60% annual increases in data center revenue and plans the MI450 accelerator for H2 next year, positioning AMD to capture share from Nvidia. Together the results underscore accelerating demand for AI infrastructure and the potential for material re-rating of both names, though CoreWeave remains volatile and loss-making while AMD's valuation still reflects high growth expectations.

Analysis

Market structure: The immediate winners are GPU-focused infrastructure (CRWV) and semiconductor suppliers (NVDA, AMD) as AI workloads push specialized cloud and accelerator demand up ~30%+ CAGR through 2033. Legacy hyperscalers (AMZN, MSFT) face margin pressure or must outspend to match specialization, increasing pricing power for AI-optimized providers and keeping GPU spot/pricing tight vs. supply — expect elevated GPU ASPs and sustained elevated option implied vols for 6–12 months.

Risk assessment: Tail risks include export controls/antitrust on NVDA/AMD, a sudden capital markets squeeze that impairs loss-making CRWV given its $1.9B liquidity (watch cash < $1B), and MI450 execution failure. Timeframes: immediate (days) — elevated vol and headline sensitivity; short (weeks–months) — MI450 marketing/tests and CoreWeave quarterly cash burn; long (quarters–years) — data-center revenue mix shift and potential winner-take-most dynamics.

Trade implications: Favor conviction in AMD execution: a 2–3% long core position via cash or 12–18 month LEAPS, scale to 5% if data-center revenue >50% or MI450 benchmarks show parity within 3 months post-launch. Use a small tactical short/speculative CRWV position (0.5–1%) via 3–6 month puts if burn >$150M/q or liquidity falls 20% post-quarter. Consider AMD/CRWV pair: long AMD (3%) funded by short CRWV (1%) to express capture of share gains while limiting macro beta.

Contrarian angles: Consensus underestimates the chance AMD compresses NVDA’s premium if MI450 hits price/perf parity — that would re-rate AMD (potentially +50% over 12–24 months) while causing NVDA multiple compression. Conversely, CoreWeave could be an acquisition target (MSFT/AMZN) if cash burns justify buyout, so avoid overly large naked shorts; monitor ASPs, gross margins, and customer concentration as leading indicators.

Key Decisions for Investors

• Establish a 2–3% portfolio long in AMD (NASDAQ: AMD) via cash or buy 12–18 month LEAPS to capture MI450 upside; increase to 5% if data-center revenue >50% or MI450 independent benchmarks show <=10% perf gap vs. NVDA within 3 months of launch.
• Initiate a tactical 0.5–1% short exposure to CoreWeave (NASDAQ: CRWV) using 3–6 month put options (strike ~15–20% below current price) to hedge funding/runway risk; trim if CRWV liquidity remains >$1.5B and quarterly loss improvement continues below $75M.
• Execute a relative-value pair: long AMD 3% / short CRWV 1% (stock or options) to express secular share shift to chip vendors versus specialist cloud providers; rebalance after MI450 launch or CoreWeave quarterly report.
• If implied volatility on AMD spikes >40% around MI450 events, switch to defined-risk 9–12 month call spreads (buy low strike, sell ~25% higher strike) sized at 1–2% of portfolio to cap downside while retaining upside exposure.
• Avoid large naked short on NVDA; instead if seeking downside exposure, buy 1–2% of portfolio in long-dated (9–12 month) put spreads on NVDA keyed to a >20% drawdown trigger, and monitor regulatory/export-control headlines over next 90 days as immediate catalysts.