Intel’s first data center GPU introduced

Intel’s first data center GPU introduced

Intel announced that it will release the Intel oneAPI software toolkit in December, which is part of Intel’s integrated approach to hardware and software design. In addition, Intel also introduced its first discrete graphics processing unit (GPU) for the data center, the Intel Server GPU, based on the Xe-LP microarchitecture and designed specifically for cloud gaming and media streaming for Android.

Intel oneAPI tools take full advantage of the latest hardware capabilities and instructions such as Intel AVX-512 and Intel DL Boost for processors, as well as features unique to the XPU. They are built with proven and proven Intel developer tools. Intel says oneAPI represents “a bold vision of a unified and simplified cross-architectural programming model that delivers uncompromising performance without being tied to proprietary technologies and enables legacy code to be integrated.”

A free version of Intel oneAPI will be available (on-premises and in Intel DevCloud), as well as commercial versions that include Intel technical support. Intel is migrating its Intel Parallel Studio XE and Intel System Studio products to oneAPI.

As for the Intel Server GPU, this 96 CU processor is used in the H3C XG310 PCIe Gen3 x16 expansion card. Its configuration includes four processors, each of which is linked by a 128-bit bus with 8 GB of LPDDR4X memory. Intel estimates that a system with two such cards can support 100 simultaneous Android users in cloud gaming. Shipping of these GPUs has already begun.

Intel’s first data center GPU introduced

Leave a Comment