Developer Tools and Software for Intel® Data Center GPU Flex Series
Review use cases, get started topics, tools, and libraries that are optimized for Intel® Data Center GPU Flex Series hardware—the industry's most open GPU solution for the intelligent visual cloud.
Virtual Desktop Infrastructure (VDI)
Remote workers can access the applications and data they need, when and where they need it, using the VDI support in Intel Data Center GPU Flex Series. Additional support includes virtualization through Passthrough and Single Root I/O Virtualization (SR-IOV) approaches.
- Passthrough is based on Intel® Virtualization Technology (Intel® VT) for Directed I/O (Intel® VT-d). This approach allows direct assignment of an entire GPU to a single user, passing native driver capabilities through the hypervisor without any limitations.
- SR-IOV allows a device, such as a GPU card or network adapter, to separate access to its resources among various PCIe* hardware functions. This "fractionalizing" of the GPU allows separate workloads or users to share the GPU resources simultaneously.
Using Intel Data Center GPU Flex Series for VDI has the following advantages:
- Low total cost of ownership (TCO) option
- Hardware-based virtualization via SR-IOV and PCIe passthrough
- Configurable dynamic scheduler
- No virtualization license fees
- No support calls for licensing server issues
Get Started
Try the following validated, publicly available VDI solutions:
Supported remoting agents:
- Citrix* Virtual Apps and Desktops 7 v23.11+
- VMware Horizon* 8 v23.03+
- GuestOS – WinClient 10/11 and Windows Server* 2019/2022
Enable SR-IOV on the following hardware:
More Resources
- Watch: Accelerate VDI with Intel Data Center GPU Flex Series
- Watch: Intel Data Center GPU Flex Series 170—Subscription-Free GPU Accelerated VDI on Proxmox* 8.1
- Intel Data Center GPU Flex Series for VDI Solution Brief
- VDI Evolution Accelerated: Intel Data Center GPU Flex Series Takes the Virtual Reins
- Overview and Performance of VDI on Intel Data Center GPU Flex Series
- Boost Video and Graphics Delivery in Horizon with VMware Blast and AV1 Enabled on Intel GPUs
- Cisco UCS X-Series with Intel Flex Data Center GPUs and VDI
- Accelerate Remote Desktops with Intel Data Center GPU Flex Series
- VMware Compatibility Guide
AI Inference
AI is transforming how we live and work every day. You can use Intel Data Center GPU Flex Series for a range of inference use cases. This includes large language models (LLM) as well as visual inference scenarios with image input, such as healthcare, smart cities, and frictionless retail applications. AI Tools provide optimized extensions for AI frameworks like TensorFlow* and PyTorch*. Optimize and deploy AI inference with the Intel® Distribution of OpenVINO™ toolkit. Combine with hardware accelerated decode, encode, and frame processing from Intel-enabled open source FFmpeg* and GStreamer industry-standard media frameworks to build pipelines with great end-to-end performance.
Intel Data Center GPU Flex Series supports popular AI frameworks:
- TensorFlow Optimizations from Intel
- Intel® Extension for TensorFlow* on GitHub*
- PyTorch Optimizations from Intel
- Intel® Extension for PyTorch* on GitHub
- Intel Distribution of OpenVINO Toolkit
- Get Started with the OpenVINO Toolkit
For hardware-accelerated decode, encode, and frame processing, use Intel-optimized FFmpeg and GStreamer or Intel® Video Processing Library (Intel® VPL).
Construct Visual Inference Pipelines
To build your own visual inference pipeline, you can use the following components:
Hardware accelerated media:
- Codecs: JPEG, AVC, HEVC, VP9, AV1
- Frame processing: resize, crop, color space conversion
Model support:
- Optimized: SSD MobileNet*, ResNet*-50, YOLOv4*, EfficientNet-B0, EfficientNet-B3, Mask-RCNN, and more
- Over 300 models functionally validated
Get Started
Take advantage of support for a wide range of models via frameworks like OpenVINO, Intel Extension for PyTorch, Intel Extension for TensorFlow, Intel® XPU back end for Triton, and more.
Intel® AI Reference Models include Linux* containers to quickly replicate the complete software environment that demonstrates the best-known performance of each of these target model or dataset combinations.
Media Delivery
With Intel® Data Center GPU Flex Series, you get a general-purpose data center graphics processor with compute and video processing capabilities. The media capabilities can be used for virtual desktop infrastructure (VDI) and visual inference, but they can also be used on their own for use cases like media transcode and streaming.
Access Intel hardware-accelerated video decode, encode, and frame-processing features via FFmpeg* and GStreamer. These industry standard frameworks allow video applications to "just work" across accelerators, hardware generations, and vendors with minimal code changes, making the oneAPI vision a reality for media too.
Get Started
Start by using media capabilities with FFmpeg and GStreamer:
Industry-Standard AI and Media Framework Support Provides Portability
Accelerate time-to-market for inference, media, and virtual desktop infrastructure (VDI) with Intel's fully open software stacks. Multiple AI extension projects from Intel allow applications based on industry-standard AI frameworks (like PyTorch* and TensorFlow*) to run on Intel® Data Center GPU Flex Series hardware with minimal changes. Industry-standard media frameworks like FFmpeg* and GStreamer are also enabled to access Intel Data Center GPU Flex Series video decode, encode, and frame-processing hardware acceleration capabilities.
Start with the Intel® oneAPI Base Toolkit for heterogeneous development across CPUs, GPUs, and FPGAs. It is open source, based on open standards, and features an industry-leading C++ compiler that implements SYCL*—an evolution of C++ for heterogeneous computing. A range of performance libraries provide portable acceleration. Enhanced performance profiling, design assistance, and debug tools are also included.
Access low-level media and compute capabilities via oneAPI Level Zero, OpenCL™ APIs, Vulkan*, and Intel® Video Processing Library (Intel® VPL).
More Resources
Get Started Guides & Articles
Intel® Tiber™ AI Cloud
Intel® Data Center GPU Flex Series Is Available Now in the Intel® Tiber™ AI Cloud
Get what you need to build, test, and optimize your oneAPI projects for free.
Product and Performance Information
† For workloads and configurations, see the Intel Data Center GPU Flex Series Performance Index. Results may vary.