"Intel announced a new data center GPU codenamed Crescent Island at the OCP Global Summit 2025. The chip is intended for AI inference applications. It combines higher memory capacity and energy efficiency in a design suitable for air-cooled servers. Crescent Island is part of Intel's renewed effort to gain a foothold in the AI accelerator market. In recent years, the company has lost market share to competitors such as Nvidia and AMD."
"From a technical standpoint, the choice of LPDDR5X memory is noteworthy. Crescent Island has 160 gigabytes of memory. It uses LPDDR5X instead of high-bandwidth memory (HBM), which is common in Nvidia and AMD GPUs. HBM has become scarce and expensive due to rising demand in the AI sector. Intel's choice of LPDDR5X therefore seems like a practical way to reduce costs and simplify production, though it may come at the expense of peak bandwidth."
Intel introduced Crescent Island, a data-center GPU designed for AI inference with a focus on higher memory capacity and energy efficiency for air-cooled servers. The GPU uses the new Xe3P architecture, a development of the Xe3 generation optimized for energy and cost efficiency. Crescent Island includes 160 gigabytes of LPDDR5X memory instead of costly high-bandwidth memory, trading peak bandwidth for lower cost and simpler production. The GPU targets workloads such as tokens-as-a-service and generative AI. An open software stack is in development and currently being tested on Arc Pro B-series GPUs. First units are expected in the second half of 2026, with Intel planning annual GPU generation updates.
Read at Techzine Global
Unable to calculate read time
Collection
[
|
...
]