Intel has enhanced cloud gaming and AI workloads for Windows with continuous software updates as the adoption of Flex Series GPUs increases.

Intel announced on the 22nd that it has expanded the production-grade software capabilities of its Intel Flex Series GPUs, including new support for cloud gaming, AI inference, and digital content creation for Windows.

Intel® Data Center GPU Flex Series has introduced flexible general-purpose graphics processing units (GPUs) for data centers and intelligent visual clouds.

As the industry increasingly adopts Flex Series GPUs, customers, solution providers, and developers are utilizing the hardware's capabilities in various ways.

For example, cloud service providers are deploying cloud game instances for Windows with streaming and media acceleration, and Media Studio is testing and deploying Flex Series GPUs for high-density streaming and transcoding.

In addition, digital content creators are using the platform for real-time rendering utilizing ray tracing hardware acceleration.

In particular, the Flex series GPUs have added cloud gaming for Windows to keep pace with the growth of the gaming market. Intel's new software features for Intel Flex Series GPUs, combined with 3rd and 4th Gen Intel Xeon Scalable processors, enable the realization of new capabilities and benefits across real-world workloads.

It provides excellent performance even in AI workloads.

In AI inference, Intel Flex Series GPUs support most common AI frameworks, such as TensorFlow and PyTorch, through the Intel® AI Analytics Toolkit. Intel has completed the validation of more than 100 AI inference models.

AI Visual Inference includes the open-source GStreamer multimedia framework, the Intel® OpenVINO™ toolkit inference engine, and open-source sample implementations of media analysis frameworks for creating complex visual inference and analysis pipelines.

In autonomous driving (AD), it demonstrates better performance than data center competitors in CARLA benchmark performance utilizing the open source solution Unreal Engine 4.

In a single 1080p sensor scenario, a single Intel® Data Center Flex 170 GPU achieves 56 frames per second (fps), which is 16% faster than the NVIDIA A10G (48fps), 40% faster than the NVIDIA GPU Tesla T4 (40fps), and 60% faster than the AMD Radeon Pro V520 MxGPU (35fps).

The performance of the Intel Data Center Flex 170 GPU scales according to four 1080p sensors, reaching 23fps, while the Nvidia A10G records 19fps.

In relation to digital content creation, real-time rendering is provided through ray tracing hardware acceleration on the Intel Flex Series via Intel® Embree.

Using the latest Intel® Open Image Denoise, AI-based noise removal can be completed in milliseconds. Productivity can be further increased by using a single SYCL codebase for CPU/GPU rendering through a single API.

The Flex Series provides customers with comprehensive graphics solutions, an open full software stack, free licensing costs, and an integrated programming model for CPUs and GPUs, delivering performance and productivity through a single API. Software stacks and development tools supporting workloads can be found in the software for Intel Data Center GPU Flex Series.

System manufacturers are preparing more than 70 designs featuring Flex Series GPUs, and more than 40 designs are scheduled to be released by the middle of this year.

Cisco, Dell, H3C, HPE, IEIT, Lenovo, Nettrix, PowerLeader, Supermicro, xFusion, and others are launching products equipped with Intel Flex Series GPUs.

The latest updates on system design, release timing, and product updates can be found on the Intel Data Center GPU Flex Series page at Intel.com.