Intel announced a key addition to its AI accelerator portfolio, a new Intel Data Center GPU code-named Crescent Island. Accordingly, the soon-to-be-released GPU will be able to meet growing demands of AI inference workloads and will offer high memory capacity and energy-efficient performance.

Sachin Katti, CTO of Intel said on Oct. 14 at the 2025 OCP Global Summit in San Jose, California that AI is shifting from static training to real-time, everywhere inference driven by agentic AI.

“Scaling these complex workloads requires heterogeneous systems that match the right silicon to the right task, powered by an open software stack. Intel’s Xe architecture data center GPU will provide the efficient headroom customers need —and more value—as token volumes surge.” 

Optimized for Inference Workloads

Because inference becomes the dominant AI workload, Intel said success depends on systems-level innovation, more than powerful chips. For that reason, Intel has worked with communities like Open Compute Project (OCP) to enable AI inference to run everywhere it is needed the most.

With Crescent Island, Intel expects it to be power- and cost-optimized for air-cooled enterprise servers and to incorporate large amounts of memory capacity and bandwidth, optimized for inference workflows.

Among its key features include Xe3P microarchitecture with optimized performance-per-watt; 160GB of LPDDR5X memory, and a support for a broad range of data types.

Boosts AI Strength

The announcement on Intel data center GPU code-named Cresent Island came a week after the company confirmed the architectural details of the company’s Intel Core Ultra series 3, also known as Panther Lake, and is built on 18A, Intel’s most advanced semiconductor process.

The Panther Lake is set to power broad spectrum of consumer and commercial AI PCs, gaming devices, and edge solution. It is the first client system-on-chips (SoC) built on Intel 18A. It offers up to 16 new performance-cores (P-cores) and efficient-cores (E-cores) delivering more than 50% faster CPU performance vs. previous generation. The new Intel Arc GPU delivers more than 50% faster graphics performance than its previous generation, and a next-level AI acceleration of up to 180 Platform trillions operations per second (TOPS).

Panther Lake will enter high volume production at Intel’s newest fab in Arizona, Untied States later this year.

15 October 2025