U.S. chipmaker Intel is returning to the GPU market: it has unveiled a new lineup, the Intel GPU D-Series. The GPUs are designed for inference—running AI models that have already been trained. The new chips are expected to intensify the company’s competition with Nvidia, AMD and Qualcomm, Yahoo Finance reports.
Details
Unlike chips for server racks, such as Nvidia Blackwell, Intel’s D-Series GPUs will be shipped as PCIe cards. They look similar to desktop gaming graphics cards. Intel said it is working with a number of hardware manufacturers, contract developers and system manufacturers, as well as neocloud companies and other partners, to bring D-Series-based systems to market. These include ASUS, HPE, Dell and Supermicro.
Before the D-Series, Intel’s last full-fledged server GPU for AI and high-performance computing was the Data Center GPU Max series, released in 2023. Intel planned to replace it with two projects—Rialto Bridge and Falcon Shores—but neither ultimately entered production. At the same time, the company was developing its specialized Gaudi AI accelerators, introduced in 2024, which were intended to compete with Nvidia’s H100.
Intel shares fell 3.2% at the close of trading on October 12. But since the start of 2026, the stock has risen 183%. Investor interest has been driven primarily by the company’s CPUs. They were overshadowed in the early years of the generative AI boom, but their role is growing again as AI agents become more widespread, particularly for AI service providers and data center operators, Yahoo Finance notes.
Context
Competition in the data center processor market is intensifying: Intel’s longtime rival AMD is rapidly gaining market share in this segment. Qualcomm has also entered the fray, bringing its own Arm-based processors, AI accelerators, memory solutions, and networking infrastructure to market as it seeks to expand beyond smartphones and reduce its dependence on that market. In addition, major cloud service providers are developing their own AI accelerators for data centers. Google, Amazon, Meta, and Microsoft have such solutions, as do Alibaba and Baidu in China. ㅤ