Why Intel Stock Just Popped

Wells Fargo raised its AMD price target to $615, supporting an AI thesis that AI spend is shifting from GPU training toward CPU inference. The note forecasts AMD GPU revenue of $15.6B this year rising to $63B by 2028, alongside CPU revenue growth of +68% in 2026, +28% in 2027, and +22% in 2028. AMD shares were up about +4.2% through 11 a.m. ET, while the article flags that Intel may benefit if the CPU narrative plays out.

Analysis

This is better read as a relative-value signal than a clean AI semiconductor repricing. If inference economics do migrate toward CPUs, the first-order beneficiaries are server CPU vendors with installed-base leverage and platform pricing power; the second-order winner is whoever can sell the whole data-center bill of materials, not just the compute chip. That makes AMD the cleaner fundamental lever than Intel, while Intel’s move is more likely a flow/short-covering reaction than an immediate earnings revision.

The near-term risk is that the market extrapolates one analyst view into a structural share shift before any of it shows up in reported revenue. CPU mix improvements typically take a couple of quarters to impact revenue, and 6-18 months for margin leverage to become visible; by then, execution risk, capex burden, and competitive pricing can easily offset the story. If hyperscaler budgets stay weighted toward training, the whole CPU-inference thesis loses steam and NVDA reasserts leadership.

Contrarian take: the consensus may be overestimating substitution and underestimating complementarity. Inference workloads often scale with memory bandwidth, networking, and software efficiency, so a ‘CPU wins, GPU loses’ narrative may be too linear; the real upside may accrue to whichever vendor captures the control plane around the accelerator ecosystem. That suggests today’s relative strength in INTC is fragile, while AMD’s upside is more durable but already partially discounted.

Key Decisions for Investors

• Relative-value trade: long AMD / short INTC for 1-3 months, targeting further dispersion if the market keeps pricing CPU-inference share gains; stop if AMD underperforms SOXX for two consecutive weeks or if next hyperscaler capex updates remain GPU-led.
• Do not chase the INTC pop outright; fade strength with a small starter short or put spread only after the initial squeeze cools, since the move looks more flow-driven than fundamentals-driven.
• Hold NVDA as the structural hedge rather than shorting it on this headline; if the next earnings cycle shows inference still GPU-heavy, NVDA should regain leadership quickly.
• Set an alert for AMD data-center CPU growth commentary over the next 1-2 quarters; if management does not confirm inflection, the CPU narrative is likely overextended and the AMD/INTC pair should be reduced.