Nvidia forecasts $108B in quarterly revenue, assumes zero chip sales to China

Nvidia just posted $96.2 billion in quarterly revenue and told Wall Street to expect even more next quarter. The $108 billion Q3 guidance comes with a remarkable asterisk: the company baked in exactly zero dollars from data center chip sales to China.

The numbers behind the confidence

Nvidia’s fiscal Q2 2027 results show a company that has more than doubled its revenue year-over-year, with growth clocking in at 106%. Data center revenue alone hit $89 billion for the quarter, making up the vast majority of the top line.

China, once responsible for somewhere between 13% and 20% of Nvidia’s data center revenue, contributed less than 1% this time around. A small batch of H200 chips went to roughly 10 approved Chinese firms, but uptake was minimal due to local purchasing restrictions layered on top of US export controls.

Total revenue from China across all product categories, not just data centers, landed at approximately $7.88 billion. That figure includes gaming GPUs, automotive chips, and other non-restricted products. For context, the data center segment alone generated more than 11 times that amount.

The Q3 guidance of $108 billion assumes this near-zero Chinese data center contribution continues. Despite that assumption, Nvidia projects about 70% revenue growth for the full fiscal year 2028.

How China went from core market to rounding error

The transformation happened in stages. US export controls first restricted Nvidia’s most advanced chips from being sold to Chinese customers. Nvidia responded by designing downgraded versions specifically for the Chinese market, but subsequent rounds of restrictions closed those workarounds too.

Then came restrictions from the Chinese side. Beijing imposed its own limitations on purchasing certain US-made chips, creating a policy sandwich that squeezed Nvidia out from both directions. The result: a market that analysts estimate could be worth $50 billion annually in AI chip demand is effectively off-limits.

CEO Jensen Huang has acknowledged the situation publicly, framing it as a temporary constraint rather than a permanent loss. The company’s financial planning, however, treats the revenue as gone until proven otherwise.

The summit variable

There’s a wildcard sitting on the geopolitical calendar. A potential Trump-Xi summit could reshape the export control landscape, and with it, Nvidia’s addressable market. Any easing of chip restrictions would instantly reopen a revenue channel that Nvidia has conservatively valued at zero.

If China’s AI chip demand truly sits near $50 billion annually and Nvidia could capture even a fraction of that, the revenue impact would be significant. Every percentage point of that market recaptured would represent billions in incremental sales.

What this means for markets and the AI supply chain

Nvidia’s ability to guide for $108 billion without China sends a clear signal about AI infrastructure spending in the rest of the world. Hyperscalers like Microsoft, Amazon, Google, and Meta are pouring capital into data centers at a pace that apparently compensates for an entire country’s worth of lost demand.

For semiconductor investors, the China dynamic creates an unusual situation. Nvidia’s current valuation reflects a world where Chinese sales are essentially zero. Any policy shift that reopens that market becomes pure upside, not a recovery to baseline.

Chinese AI firms, meanwhile, have been forced to accelerate development of domestic alternatives. Huawei’s Ascend chips and other local solutions are gaining traction, though they remain generations behind Nvidia’s architecture in most benchmarks. The longer export controls persist, the more entrenched these alternatives become, potentially shrinking Nvidia’s addressable market even if restrictions eventually ease.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.