Nvidia's Forecast Assumes No Data Center Chip Sales to China. The Sept. 24 U.S.-China Summit Could Change That.

Nvidia forecast fiscal Q3 revenue of $108.0 billion, up roughly 89% year over year from $57.0 billion, while assuming zero data-center compute revenue from China. Fiscal Q2 revenue rose 106% to $96.2 billion, including $89.0 billion in data-center sales, demonstrating that its growth outlook currently does not depend on reopening the Chinese AI-chip market. A potential U.S.-China summit breakthrough on chip restrictions could add upside—historical China data-center demand could be worth nearly $20 billion per quarter—but export controls, Chinese policy, tariffs and local competition remain significant obstacles.

Analysis

China optionality should not be capitalized at face value. Any bilateral accommodation is more likely to permit controlled, lower-performance products with tariffs, compliance friction, and local-purchase constraints than restore Nvidia's historical economics; that outcome adds revenue but can be margin-dilutive and risks diverting scarce advanced packaging capacity from higher-return non-China shipments. The more material near-term signal is whether management can sustain ex-China sequential growth without extending receivables, increasing inventory, or reducing gross-margin guidance.

A durable exclusion from China strengthens the local AI hardware stack rather than simply leaving demand unserved. Huawei, Cambricon, SMIC and domestic server integrators gain software, qualification and installed-base advantages that compound over 6-18 months; once workloads are ported, a later reopening may recover less share than investors assume. TSMC remains comparatively insulated because global hyperscaler demand, not China-bound Nvidia volume, is the bottleneck, while AMD and Broadcom have limited direct upside from an Nvidia-specific licensing thaw and could face a relative multiple headwind if Nvidia receives a tradable China catalyst.

The market may be underpricing regulatory asymmetry: Washington can tighten rules quickly, while Beijing can suppress procurement even after U.S. licenses are granted. Thus a summit headline can produce a days-long NVDA rally, but a sustainable rerating requires verifiable purchase orders, product approval, and evidence that economics clear tariff and inspection costs. Conversely, a disappointment should not impair the base case if the next earnings release confirms backlog conversion and stable gross margin; the key falsifier is a downward revision to ex-China data-center revenue or a material increase in channel inventory.

Key Decisions for Investors

• Maintain or initiate a 1-3 month long NVDA position only against a defined catalyst: add on post-summit weakness if management or supply-chain checks continue to support ex-China demand, rather than buying a diplomatic headline. Risk/reward is favorable only if the position is underwritten on the existing outlook; exit or reduce if next-quarter gross-margin guidance falls materially or ex-China growth decelerates below the implied sequential ramp.
• Express China-opening optionality with a limited-risk NVDA call spread expiring after the next earnings print, not outright short-dated calls around the summit. A modest license or procurement announcement can lift the stock, but full-market reopening is improbable; capped upside avoids paying for an unrealistic normalization scenario.
• Pair long NVDA / short AMD over the next 1-3 months on a modest, beta-neutral basis. Nvidia retains the cleaner beneficiary profile from any approved China shipment and stronger visibility into high-end AI demand, while AMD has less direct China upside; cover if AMD shows material accelerator share gains, a major hyperscaler design win, or Nvidia's gross-margin outlook weakens.
• Do not add China semiconductor proxies solely on an export-control thaw. Instead, monitor Huawei/Cambricon procurement and SMIC capacity signals over 6-18 months; sustained domestic substitution is the structural trade, whereas a limited Nvidia license regime may reinforce Beijing's incentive to localize rather than reverse it.