Optimizing Performance and Improving Uptime with NVIDIA GRID

by Raj Rao

Our latest GRID release brought a number of updates and enhancements that deliver a more robust user experience with increased system uptime for any vGPU user profile. For a deeper understanding on how to optimize your GRID deployment, check out some of the resources below.

Better security, resource management and host lifecycle management: Support for VMware’s vSphere 6.5, the latest version of the VMware virtualization platform, brings value-added features and a universal application platform for running any application.

Improved system uptime: NVIDIA GRID deployments can benefit from VMware’s HA support for vGPU. HA offers failover protection when there are system outages by restarting the affected VMs on available hosts in the same cluster. Check out our GRID and HA support demo. For more information on HA for vGPU backed VMs, consult VMware documentation for vSphere 6.5.

Greater system stability and performance: This latest release reduces framebuffer footprint for all profiles by up to 70MB, which is more than 10% of the total framebuffer available for the 0B and 0Q profiles. This optimization allows GRID to return the extra framebuffer back to the pool for user applications, improving system stability and performance, especially for configurations using smaller framebuffer profiles. For more information on optimizing profiles, see our blog post.

Closing out the year with new functionality is a testament to the GRID software model that allows us to continuously bring innovation to graphics virtualization and our customers. Looking forward, we predict great things for the virtualization space. You can read all about it in our 2017 predictions blog.