AMD Just Hit 2030's Earnings Target Two Years Early. Here's What Happens Next

AMD’s Q1 FY2026 results were strong, with revenue of $10.253 billion, Data Center revenue up 57% year over year to $5.775 billion, and non-GAAP EPS of $1.37, while Q2 guidance points to roughly $11.2 billion in revenue, about 46% growth. Bernstein’s Stacy Rasgon turned more constructive, arguing AMD’s server CPU opportunity has expanded and that 2028 earnings could approach $20, two years ahead of the prior 2030 target. The GPU opportunity remains more uncertain, with AMD paying equity warrants to win large AI deals and export controls costing about $440 million in FY2025.

Analysis

The key second-order shift is that AMD’s bull case is no longer hostage to GPUs; the server CPU engine can re-rate the entire equity even if accelerator execution stays merely “good enough.” That matters because CPU share gains are slower-moving, stickier, and less capex-intensive than GPU share grabs, which means incremental revenue can convert to profit with less dependence on ecosystem lock-in. In other words, the market may be underpricing a more durable earnings staircase from CPUs while still treating AMD as a single-thread AI beneficiary.

The warrant-heavy AI deals are the real tell: AMD is effectively subsidizing its way into relevance, which should be read as both a competitive moat-building expense and a signal that its product gap is still non-trivial. That creates a strange asymmetry: the deals can accelerate ROCm adoption and customer validation over the next 6-18 months, but they also cap near-term economic value per share because dilution arrives before software leverage does. Investors focusing only on headline design wins may be missing that the first derivative of revenue could improve faster than the second derivative of margins.

Relative positioning still looks asymmetric versus the ecosystem. Intel is the structural loser if AMD’s CPU share continues compounding, but the more interesting loser may be NVIDIA at the margin: every incremental dollar AMD spends to build developer traction is a dollar that reflects how expensive the GPU onboarding war has become. The contrarian risk is that the CPU upside is now consensus-adjacent, while the GPU optionality remains farther from monetization; if the market starts discounting the dilution and the export-control drag more heavily, the stock can de-rate even with solid execution. The next 1-2 quarters matter for validation; the next 2-3 years matter for whether AMD’s AI franchise earns its cost of customer acquisition.

Key Decisions for Investors

• Stay long AMD on a 6-12 month horizon, but size it as a self-funding momentum/quality trade rather than a pure AI option; upside comes from CPU earnings compounding, downside is multiple compression if dilution outruns operating leverage.
• Prefer a long AMD / short INTC pair for the next 2-4 quarters: the spread benefits from continued server share capture and should outperform if AMD keeps converting data-center revenue into FCF faster than Intel can stabilize.
• For accounts willing to express the GPU debate, use AMD call spreads 6-12 months out instead of outright stock; the convexity captures a successful ROCm/MI450 ramp while limiting damage if warrant dilution and execution slippage cap the rerate.
• Reduce or hedge NVDA relative exposure against AMD strength over the next 3-9 months: if AMD’s CPU story becomes the primary driver, the market may begin to price a more credible second-source AI ecosystem, which can compress NVDA’s scarcity premium at the margin.
• Watch for post-earnings pullbacks to add AMD rather than chase strength; if the multiple stays above mid-60s forward earnings, the entry point matters more than the thesis, and any disappointment on China/export or dilution can trigger a sharp de-rate.