Intel Arc B570 10GB GDDR6 PCIe 5.0 Desktop GPU

The Intel Arc B570 is a discrete desktop graphics card built on Intel's Battlemage (Xe2) GPU architecture, featuring 10GB of GDDR6 memory across a 192-bit memory interface. Powered by the Xe2-HPG microarchitecture, it delivers hardware-accelerated ray tracing, mesh shading, and XeSS (Xe Super Sampling) AI-based upscaling, making it a capable mid-range GPU for both professional visualization and compute-adjacent workloads. The card connects via a PCIe 5.0 x8 interface, ensuring forward compatibility with current-generation enterprise server and workstation platforms.

Equipped with 20 Xe2 cores (1280 shaders), the Arc B570 supports Intel's XMX (Xe Matrix Extensions) AI acceleration engines, which accelerate INT8 and FP16 matrix operations relevant to inference tasks and AI-assisted content pipelines. The card includes hardware AV1 encode and decode support, enabling efficient media transcoding at scale. Display output is provided via four Thunderbolt 4 / DisplayPort 2.1 connectors, supporting up to 4K and beyond across multiple simultaneous displays.

For enterprise IT environments requiring scalable, software-compatible GPU resources, the Arc B570 benefits from Intel's open-ecosystem driver strategy, with strong support for DirectX 12 Ultimate, Vulkan, OpenCL, and oneAPI, the latter being particularly relevant for heterogeneous compute deployments. The card operates within a 150W TDP envelope and requires a single 8-pin PCIe power connector, fitting standard ATX workstation builds without specialist power infrastructure.

Ideal for

• AI inference acceleration for edge and on-premises workloads using XMX INT8/FP16 matrix engines via Intel oneAPI
• High-density AV1 media transcoding for enterprise video streaming, digital signage, and broadcast production pipelines
• Multi-display professional visualization and CAD rendering on workstations requiring DirectX 12 Ultimate and Vulkan compliance
• Heterogeneous compute development and testing leveraging Intel oneAPI across CPU and GPU targets
• Virtual desktop infrastructure (VDI) host GPU acceleration for graphically intensive remote workloads
• Simulation and scientific visualization workloads requiring OpenCL support within cost-sensitive departmental workstation budgets

Technical specifications

Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.