AMD: The Next $1 Trillion Company

AMD reported Q2 revenue up 50% and non-GAAP EPS up 82%, driven by data center strength, with guidance above consensus. However, gross margin is expected to be flat at ~56% as HBM4 costs pressure profitability. Valuation is still rich (non-GAAP P/E 81.72; FCF yield 1.1%), but the forward earnings backdrop improved versus the recent drawdown.

Analysis

The key signal is not the top-line beat; it is that incremental AI/data-center demand is still not translating into margin leverage. That means AMD is behaving like a share-gainer in a capital-intensive phase of the cycle, which is usually positive for unit growth but less favorable for cash returns than the market expects at an 80x+ earnings multiple. In that setup, the stock can keep reacting well to revenue momentum, but the multiple becomes fragile if investors realize the earnings power is being capped by input costs rather than constrained demand.

Second-order beneficiaries are the HBM and advanced-packaging ecosystem rather than AMD itself: Micron and the broader memory stack capture pricing power when accelerator supply chains stay tight, while TSMC/packaging capacity and toolmakers benefit from continuing AI buildout. The relative loser is Intel, because sustained AMD data-center share gains pressure the legacy x86 profit pool even if AMD’s own margin expansion is muted. Compared with Nvidia, AMD still looks like the lower-quality way to own the AI infrastructure theme because it is not yet proving superior operating leverage.

Near term, the reaction can stay constructive for days, but the real catalyst window is the next 1-3 months: investors will test whether management can convert backlog into gross-margin expansion or whether this is simply high-volume, low-leverage growth. The contrarian risk is that consensus is extrapolating AI strength into a durable rerating while ignoring that the earnings curve is flattening beneath the revenue curve. The bullish thesis breaks if gross margin stays pinned in the mid-50s and FCF yield does not improve as the revenue base scales.

Key Decisions for Investors

• Prefer long Micron (MU) over AMD for the next 1-3 months: same AI supply-chain tailwind, but MU has more direct leverage to HBM tightness and less reliance on multiple expansion.
• Use strength to underweight or trim AMD rather than chase it outright; at current valuation, the stock needs visible gross-margin expansion to justify further rerating.
• Pair trade: long NVDA / short AMD over 1-3 months to express the view that both benefit from AI demand, but NVDA monetizes it more efficiently and should hold the premium multiple better.
• Set a watch item on AMD next-quarter gross margin and FCF conversion; if margin remains flat or guide implies continued HBM cost drag, that is the trigger to add to the short-side relative-value trade.
• If AMD gaps sharply higher on the print, consider a 4-8 week bearish put-spread hedge rather than an outright short, because the immediate tape can stay strong even when the fundamental margin story is mediocre.