1 Wall Street Analyst Just Upgraded AMD Stock and Estimates 30% Upside. Here's Why He's Bullish on This Semiconductor Stock.

Raymond James raised AMD to a strong buy with a $641 price target (~33% upside). The bullish thesis is that agentic AI will accelerate server CPU demand, with the server CPU market projected to reach about $201B by 2030 and AMD’s server CPU TAM revised to >$220B. AMD reported $6.7B in data center revenue (+107% YoY), and management guided to server revenue growth of >80% YoY in 2H26 and >70% for full-year 2027 (with data center revenue expected to more than double in 2027), though the stock trades at a pricey ~116x trailing P/E vs ~34x for the tech sector.

Analysis

The clean read-through is not “AI is good for semis,” but that server compute is likely to become more socket-diverse, which shifts bargaining power toward vendors with credible CPU share gains. That is favorable for AMD’s mix and operating leverage, but the first derivative trade is probably against Intel, where even modest share loss in a high-margin franchise matters more than the headline revenue number suggests. ARM is a quieter secondary beneficiary if custom silicon proliferates, though its monetization lags actual deployment, so the equity reaction should be less immediate.

The near-term risk is that this is mostly a multiple event until the next few prints prove conversion into booked revenue. AMD already trades like a growth victor, so any evidence of slower server attach rates, weaker gross margin expansion, or hyperscaler capex rebalancing can compress the stock faster than the underlying fundamentals would justify. The key falsifier over the next 1-3 months is not whether TAM is large, but whether management can keep raising the revenue ramp without leaning on increasingly optimistic long-dated assumptions.

Contrarian view: the market may be over-assigning linearity to a TAM estimate that could be partly circular. If more of the AI stack migrates to custom in-house silicon, AMD can still win share while the dollar pool per rack gets squeezed, which is less bullish than the headline implies. Over 6-18 months, the better expression may be relative value versus Intel rather than an outright chase in AMD; the long thesis works only if AMD’s revenue growth remains above the market’s already elevated expectations and the premium multiple is allowed to persist.

Key Decisions for Investors

• Long AMD on a 3-5% post-upgrade pullback; use a 6-12 month horizon and target 15-20% upside, but cut risk if data center growth or gross margin commentary softens on the next earnings call.
• Pair trade: long AMD / short INTC over the next 1-3 months to isolate server CPU share gains; this is the cleaner expression if the thesis is share transfer rather than broad sector beta.
• If adding upside convexity, prefer AMD call spreads with 6-12 months to expiry instead of outright stock; valuation already prices execution, so defined-risk upside is better than paying full premium.
• Watch GOOGL and META capex disclosures for custom-CPU commentary; if in-house silicon adoption accelerates faster than merchant CPU demand, that is the main thesis breaker for AMD.
• Avoid shorting NVDA as the primary hedge; the read-through is more additive compute than direct GPU substitution, so NVDA is a weaker offset than INTC.