Home > Servers > Rack and Tower Servers > Intel > White Papers > Boost VDI Performance up to 123% with Intel® Datacenter GPU Flex Series 140 on Dell PowerEdge Servers > Intel GPU Flex Series 140
Intel Data Center GPU Flex 140 is a 75-watt low profile PCIe Gen4 GPU card for accelerating VDI, cloud gaming, media processing & delivery, and AI Visual Inference applications in data center servers. Each PCIe card has 2 GPUs, each with eight Xe cores, two media engines, and 6GB of GDDR6 memory attached. Considering a graphics local memory provisioning granularity of ~1GB per virtual GPU, each Flex Series 140 GPU can support up to 12 VDI sessions with low-to-moderate graphics performance requirement per display pixel, making it a compelling solution for high density VDI deployment.
VDI encompasses a wide range of hosting configurations, user profiles, use cases, and applications. In each VM, the vGPU serves all the 3D, media, and compute workloads and encodes the framebuffers of the VDI session, thereby freeing up vCPU resources. This offload of CPU utilization towards GPU resources helps provide the resource headroom to support other desktop applications that demand higher graphics performance. Freeing up CPU resources creates more opportunities for VM density improvement, that is, more sharing of vCPUs across VMs or reducing vCPU allocation for each VM.
The results highlighted in this white paper demonstrate that Intel Data Center GPU Flex Series delivers reliable performance for VDI use cases, especially for the high-density knowledge worker segment and also for the graphics segment. With Intel Data Center GPU Flex Series, customers can experience consistent framerates, low render, and encode latencies at high quality of graphics . Each VDI session with a vGPU demonstrates sufficient GPU resource headroom, substantially reduced CPU utilization, thus improving VM density, the potential Total Cost of Ownership, and the scalability of each VDI server.