Why Nvidia Held Up While AMD and Intel Fell in the Chip Selloff

The chip selloff did not hit every company equally. In a Friday morning snapshot reported by 24/7 Wall St., Intel and AMD were each down about 3%, the iShares Semiconductor ETF was down 2%, and the Nasdaq-100 tracking QQQ fund was off 0.9%. Nvidia, by contrast, was slightly higher.

One morning does not establish a lasting trend. The split still offers a useful view of what investors are pricing: Nvidia as the incumbent AI platform, AMD as the fast-growing challenger, and Intel as a recovery story carrying the cost of its own factories.

AMD's growth is real, but so is its dependence on AI demand

AMD reported second-quarter Data Center revenue of $6.7 billion, up 107% from a year earlier. Demand for EPYC processors and Instinct accelerators drove the increase, and non-GAAP gross margin reached 56%.

That is a stronger financial base than AMD had a year ago. It also leaves the company more exposed to any slowdown in accelerator deployments. Its asset-light model means outside foundries shoulder the cost of building leading-edge fabrication plants, but AMD still depends on customers maintaining large AI infrastructure budgets.

The market can therefore reward AMD quickly when deployment forecasts rise and punish it just as quickly when financing or utilization comes into doubt.

Intel carries a different kind of risk

Intel's second-quarter results showed its strongest revenue growth in more than 15 years, according to Chief Executive Lip-Bu Tan. The improvement matters, but the company is simultaneously rebuilding its product portfolio and trying to establish Intel Foundry as a credible manufacturing business.

Intel Foundry lost $2.1 billion in the quarter. The unit is advancing the Intel 18A process and preparing 18A-P, but most foundry activity still supports Intel's own products. External customers must eventually fill those factories if the investment is to earn an acceptable return.

That makes Intel unusually sensitive to capital intensity. A weaker AI spending cycle can reduce demand for server chips at the same time that factory depreciation and development costs continue.

Why Nvidia traded differently

Nvidia enters the same debate from a position of greater operating leverage. Its fiscal second-quarter Data Center revenue reached $89.0 billion, and the company sells a broad stack spanning accelerators, networking, systems and software. Customers that standardize on that stack face meaningful switching costs.

Nvidia is not insulated from AI financing risk. Its latest filing shows large commitments to supply, cloud services and infrastructure, and it has become more involved in helping customers secure capacity. The difference is that its current revenue and margins provide more room to absorb volatility.

The three stocks therefore represent distinct exposures rather than interchangeable AI trades. AMD offers rapid accelerator growth without owning fabrication plants. Intel offers a turnaround plus a domestic foundry option, with heavy fixed costs. Nvidia offers the strongest platform economics, alongside increasing ecosystem financing exposure.

The next useful signals will come from customer spending plans, accelerator deployment schedules and Intel Foundry's path toward outside revenue. If AI budgets keep expanding, all three can participate. If financing tightens, the market is likely to keep distinguishing between platform strength, product momentum and the burden of factory ownership.