Is AMD a Top Artificial Intelligence (AI) Stock to Buy?

AMD's latest quarter showed strong AI-related momentum, with data center revenue up 57% year over year and client, gaming, and embedded revenue up 23% and 6%, respectively. However, the article argues the stock is expensive at 147x trailing earnings, 61x forward earnings, and 35x 2027 earnings, especially versus Nvidia's 26x forward earnings. Overall, the piece is constructive on AMD's business but bearish on valuation relative to the market leader.

Analysis

The market is implicitly pricing AMD as a scaled AI winner, but the more important second-order issue is that its earnings power is still being normalized off a mixed business mix, not a pure AI flywheel. That makes the multiple vulnerable if data-center growth slows even modestly: when a company’s headline growth is driven by one segment while the rest of the stack is merely adequate, the market tends to pay peak optimism prices for cyclical earnings quality. In that setup, any deceleration in GPU attach rates or margin expansion can compress the stock faster than the revenue print would suggest.

NVDA remains the cleaner expression of AI demand because it has both pricing power and ecosystem control, which matters more than revenue growth alone. The key second-order effect is supply-chain leverage: if capital spending shifts toward the leader, ODMs, advanced packaging, networking, and memory vendors tied to the dominant platform may keep compounding even if AMD continues to take share in niche pockets. AMD can still work, but the burden of proof is now on sustained earnings revisions, not just top-line acceleration.

The contrarian angle is that AMD may be less expensive than it looks only if the market is overestimating how durable NVDA’s premium is. If AI infrastructure spend broadens beyond one vendor over the next 6-18 months, AMD’s valuation could de-rate less than feared and the stock could outperform on valuation catch-up rather than fundamentals. The risk to the bearish view is that another quarter of 50%+ data-center growth plus margin expansion would force analysts to raise 2026-27 estimates, which is where the current multiple starts to look less stretched.

Key Decisions for Investors

• Prefer NVDA over AMD as the primary AI exposure for the next 3-6 months; the cleaner earnings quality and lower forward multiple offer better risk/reward unless AMD’s estimate revisions inflect materially upward.
• If holding AMD, hedge with a short AMD / long NVDA pair into the next earnings window; target 8-12% relative underperformance on any sign that AMD data-center growth is normalizing.
• Buy AMD only on a 10-15% pullback or after forward EPS estimates rise again; entry at the current multiple offers poor margin of safety if growth decelerates even slightly.
• Use call spreads on NVDA rather than outright AMD longs for AI beta exposure over the next 1-2 quarters; this captures upside while limiting multiple-compression risk.
• For more aggressive portfolios, short AMD into strength if it re-rates above prior post-earnings highs without a corresponding estimate reset; risk should be capped with a stop above the next consensus revision cycle.