Is AMD Catching Up With Nvidia? Here's the Best Stock to Buy Today.

The article argues that AMD’s rally versus Nvidia (AMD +125% YTD vs Nvidia +20% YTD, and +300% since early 2025 vs Nvidia +67%) overstates relative business momentum. It cites Nvidia’s faster scale and growth in the AI buildout—FY2027 Q1 revenue up 85% YoY to $81.6B and diluted EPS up 214%, with data-center revenue up 92%—versus AMD FY2026 Q2 revenue up 50% to $11.5B, diluted EPS up 156%, and data-center up 107%. Despite AMD growing faster in data center, the valuation gap is framed as large (AMD ~33x next FY earnings vs Nvidia ~17.6x), implying Nvidia is the better-value stock and potential leader over the next 12–18 months.

Analysis

The market is treating AMD’s faster share price move as a proxy for business momentum, but the more durable signal is valuation dispersion inside the same end-market. If NVDA sustains materially faster earnings growth while trading at roughly half AMD’s forward multiple, the next leg is likely multiple compression in AMD rather than multiple expansion in NVDA; that is a classic setup for relative-value mean reversion over the next 1-3 months, not a clean outright long in either name.

The second-order effect is on the broader AI complex: if investors start demanding proof of monetization instead of paying for optionality, the premium buckets get hit first—high-beta GPU adjacencies, AI networking, and “next-gen accelerator” narratives. That would favor the most cash-generative incumbent and pressure challengers that need continued design-win cadence to justify elevated multiples. NVDA also has an easier path to defend sentiment into earnings because its scale allows it to turn demand into guidance, whereas AMD must keep proving it can convert share gains into margin and operating leverage.

The contrarian risk is that the market is already fully aware of this spread, so the pair may not work if AMD’s next product cycle or hyperscaler diversification narrative keeps revising growth higher. In that case, the right read is not that AMD is cheap/expensive, but that the market is paying for a multi-quarter option on share capture. The thesis breaks if AMD continues >100% data-center growth with no gross-margin giveback, or if NVDA guides below the street’s implied re-acceleration at the next print.

Key Decisions for Investors

• Relative-value: long NVDA / short AMD on a 1:1 dollar-neutral basis into NVDA earnings; target 10-15% spread compression over 1-3 months if NVDA confirms stronger growth and AMD fails to re-rate higher. Falsify if AMD re-accelerates guidance or NVDA misses growth expectations.
• If you want lower event risk, wait for NVDA earnings and sell a post-print AMD bounce into strength rather than shorting weakness; the setup is better as a valuation fade than a momentum chase.
• Use NVDA call spreads instead of outright long stock if you want upside participation with less beta to the overall AI tape; this is cleaner if the thesis is relative outperformance rather than a broad sector rally.
• Watch for revision divergence in the next 30-60 days: if AMD estimates keep rising faster than NVDA while the stock still trades at a premium, stand down on the short; that would mean the market is correctly pricing share gains, not overpaying for them.