Intel CEO Lip-Bu Tan said lower-cost Chinese AI models such as DeepSeek, Kimi and GLM are gaining adoption because their performance is "good enough" for many uses, according to a Reuters report cited by Stocktwits. His argument is straightforward: cheaper inference, the process of running a trained model, can make more AI applications economical and increase the amount of computing work performed.
Why cheaper inference could help Intel
Lower model costs do not necessarily reduce total hardware demand. If a company can run an AI task at a lower cost, it may deploy the task more often or introduce it to more employees and customers. That volume effect can expand demand for server CPUs, accelerators, networking and storage even as the cost of each individual query falls.
Intel has evidence that data-center demand is already improving. The company reported second-quarter 2026 revenue of $16.1 billion, up 25% year over year. Data Center and AI revenue reached $6.3 billion, up 59%. Intel also guided third-quarter revenue to a range of $15.8 billion to $16.8 billion and non-GAAP earnings per share to $0.38.
Those figures support the compute-demand thesis, but they do not prove that open models are the cause. Intel's results also benefited from improved factory yields, execution and broader product demand. In addition, its second-quarter GAAP net loss was $11.0 billion, while non-GAAP net income was $2.2 billion, a large difference that investors should not overlook.
Open models create opportunity and competition
Open-weight models make their parameters available for developers to adapt or deploy. This can accelerate adoption because companies gain more control over cost, privacy and customization. It can also intensify competition among model providers and shift more value toward infrastructure suppliers.
Intel is positioned across CPUs, custom chips, advanced packaging and foundry services, but it still must win specific designs. A larger AI market does not automatically translate into Intel revenue if customers prefer rival processors or internal chips.
What investors should watch
The next test is whether Intel sustains Data Center and AI growth while improving GAAP profitability. Investors should compare third-quarter revenue and gross margin with guidance, track customer wins for AI servers and custom silicon, and watch whether capital spending produces better utilization rather than excess capacity.
BTI's conclusion is that lower-cost open AI models can expand the total compute market. For Intel, that is a credible demand tailwind, but execution, product competitiveness and the conversion of demand into profitable revenue remain the decisive variables.