Performance Analysis

Closely analyze the GPU frame buffer on the VM to ensure correct sizing. As mentioned in the previous section, a good rule of thumb to follow is that a VM’s frame buffer usage should not frequently exceed 90% or average over 70%. High utilization can lead to a suboptimal user experience, including degraded performance and potential crashes.

The graph below illustrates the vGPU frame buffer (FB) usage within a VM using a 2Q vGPU profile compared to a 4Q profile. In this example, the benchmark was Esri ArcGIS Pro, a professional geospatial software application and spatial navigating multi-patch 3D data. The 2Q VM reported longer rendering times and experienced software stuttering, while the 4Q VM maintained a rich and fluid end-user experience with performant render times.

Figure 11 - vGPU Framebuffer Usage within a VM

Analyzing Host resource metrics to identify potential bottlenecks when multiple VMs execute workloads is imperative for providing a quality user experience. The most successful deployments are those that balance user density (scalability) with quality user experience. Over-utilized server resources can significantly degrade user experience. The chart below illustrates host utilization rates when a benchmark test is scaled across multiple VMs, highlighting the importance of resource allocation and monitoring for optimal performance.

Figure 12 - Host Utilization Rates Across Multiple VMs

Figure 12 illustrates GPU utilization rates, indicating there is no GPU bottleneck. This suggests the server has ample headroom within the GPU compute engine. The GPU utilization time is averaged across the three L40S GPUs in the server. While GPU headroom is maintained throughout the test, CPU resources have become depleted, resulting in negatively impacted VDI performance and user experience.

Choosing the correct server CPU for virtualization and proper configuration can directly affect scalability even when a virtual GPU is present. Processor resources are often hyperthreaded and overprovisioned to a certain degree. In terms of CPU specs, consider the number of cores and clock speed. For NVIDIA RTX vWS, choose higher clock speeds over higher core counts.

An example server configuration for NVIDIA RTX vWS is provided as an appendix item.