It's About To Get Worse For Nvidia

Nvidia reported Q2 FY2027 revenue of $96.22 billion and EPS of $2.22, both ahead of consensus, and guided to 70% revenue growth in FY2028. Management cited supply constraints against roughly 100% underlying demand, underscoring sustained AI infrastructure demand. The key offset is margin pressure from rising memory costs: gross margin is expected to trough at 71%-72% in Q4 before recovering to 72%-73% in FY2028.

Analysis

The key investable change is that NVDA’s bottleneck has shifted from accelerator availability toward memory content and pricing. That transfers near-term incremental economics to HBM/DRAM suppliers—most directly MU among U.S.-listed names—while NVDA absorbs the input-cost volatility because system demand remains supply-constrained. A 100-200bp gross-margin miss would matter disproportionately for NVDA’s multiple if investors begin to view its margin structure as structurally normalizing rather than temporarily pressured.

Over the next 1-3 months, the market will likely focus less on revenue upside and more on whether memory inflation is passed through in platform pricing, especially for next-generation rack-scale systems. This is favorable for MU and potentially WDC/STX through broader enterprise-storage pricing, but adverse for AI hardware firms with lower pricing power, including smaller server OEMs and networking vendors exposed to fixed-price commitments. AMD is a secondary beneficiary only if constrained NVDA supply creates genuine workload substitution; absent evidence of customer deployments shifting, this is not yet a clean share-gain thesis.

The contrarian view is that constrained supply can conceal demand elasticity: hyperscalers may be reserving capacity ahead of actual utilization, and elevated memory costs could force procurement optimization rather than unlimited pass-through. Falsification would be a sequential deceleration in NVDA backlog conversion, a gross-margin guide below the stated trough range, or MU signaling HBM demand/pricing moderation. For the 6-18 month horizon, expanding memory capacity and alternative accelerator supply should reduce scarcity rents, making NVDA’s earnings growth durable but its valuation increasingly dependent on sustaining premium margins.

Key Decisions for Investors

• Initiate a 1-3 month pair: long MU / short NVDA in equal dollar exposure. The trade captures memory suppliers’ improving pricing power against NVDA’s input-cost and margin-risk asymmetry; exit if MU guides weaker HBM pricing or NVDA demonstrates sustained gross-margin recovery above 73%.
• Maintain NVDA core exposure but avoid adding aggressively into the immediate reaction; add only on a pullback tied to margin concerns rather than demand deterioration. Risk/reward improves if management demonstrates price pass-through or raises its gross-margin floor at the next earnings update.
• Use NVDA downside protection via 3-6 month put spreads rather than outright short exposure, targeting protection through the next margin-guidance checkpoint. The primary risk to the hedge is continued supply scarcity enabling price increases that offset memory inflation.
• Watch AMD for evidence of actual hyperscaler substitution—material accelerator revenue-guide upside, named deployment wins, or improved MI-series supply commentary. Until then, treat it as an alert rather than a long recommendation; NVDA capacity constraints alone do not establish share transfer.