Intel Data Center GPU Flex 140

The Intel Data Center GPU Flex 140 is a purpose-built enterprise graphics processing unit based on the Intel Xe-HPG microarchitecture, designed specifically for media delivery, cloud gaming, and AI inference workloads in dense data center environments. Built on a 6nm process node, the Flex 140 integrates two GPU tiles on a single card, delivering hardware-accelerated AV1, HEVC, AVC, and VP9 encode and decode capabilities that make it a compelling solution for high-density video streaming infrastructures.

The GPU is optimized for multi-user virtualization scenarios and supports Intel's media and display virtualization extensions, enabling operators to consolidate large numbers of concurrent media transcoding sessions onto a single server. Its low-profile, half-height form factor and passive or active cooling options make it suitable for 1U and 2U rack configurations where spatial efficiency and thermal management are critical operational considerations.

As part of Intel's broader Data Center GPU Flex Series, the Flex 140 is positioned for workloads that require high throughput with predictable, consistent latency rather than raw floating-point compute at scale. It supports PCIe 4.0 connectivity and is compatible with industry-standard virtualization frameworks and media software stacks including oneAPI and FFmpeg with Intel media SDK integration, allowing enterprise IT teams to integrate it into existing infrastructure toolchains with minimal re-engineering overhead.

Ideal for

• High-density video transcoding for streaming platforms requiring simultaneous AV1, HEVC, and AVC encode/decode at scale
• Cloud gaming infrastructure where multiple concurrent user sessions must be served from a single GPU card with isolated resource allocation
• AI inference at the edge or in the data center for vision and media analytics pipelines requiring hardware-accelerated pre- and post-processing
• Virtual desktop infrastructure deployments leveraging GPU-assisted graphics virtualization for knowledge workers and media production teams
• OTT and broadcast media workflows demanding real-time, low-latency transcoding with format flexibility across multiple simultaneous streams
• Telecommunications and content delivery network nodes where compute-efficient, rack-dense GPU acceleration reduces per-stream operational overhead

Technical specifications

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