I’m delighted to be able to share that the v7.2 release of eG Enterprise has added a number of significant enhancements to extend or support monitoring of GPUs in the datacenter and cloud. New GPU Monitoring capabilities in eG Enterprise v7.2 include:
- Vendor support expanded beyond NVIDIA to include AMD and Intel GPU technologies, allowing you compare and monitor on-prem GPU environments and cloud enabled GPU instances side-by-side and chose the most cost-effective and relevant options
- Extended support for NVIDIA RTX Virtual Workstation Technology in addition to our existing extensive support for NVIDIA passthrough and virtualization and sharing technologies (vGPU)
- GPU monitoring for Linux VMs and VDAs in addition to existing Windows support
- Platform support for GPU monitoring extended to enable GPU Monitoring for Nutanix AHV powered infrastructure
New Support for AMD and Intel GPUs
AMD MxGPU shared GPU technologies are now widely adopted via cloud instances offering often highly cost-effective GPU enabled options such as the G4ad instances available on Amazon AWS and the NVv4 instances available on Microsoft Azure.
With many CAD and 3D applications such as Siemens NX and Autodesk titles (AutoCAD, Revit, and Inventor) now certified for AMD-accelerated cloud technologies, we are pleased to offer our Manufacturing, Design and Media customers support as they evaluate and adopt AMD options.
In addition to AMD, we have also added support for the Intel GPU technologies which some of our customers use.
Extended Support for NVIDIA RTX GPUs
NVIDIA GPUs including the latest RTX cards can be monitored at Hypervisor, OS and VM level.
When monitoring using the ”Inside View of Desktops”, detailed diagnostics on application usage of the VM will also be available.
GPU Monitoring for Linux VMs and VDAs in Addition to Existing Windows Support
Driven largely by demand from the Media and Entertainment (M&E) and Oil & Gas simulation sectors where many rely on Linux based applications, we are pleased to add support for NVIDIA GPU technologies for Linux based digital workspaces.
Coupled with enhanced for support for NVIDIA’s RTX technologies this offers those working within applications designed for the Linux ecosystem working within the VFX framework excellent new options to ensure observability into both the performance and the cost-effectiveness of their choices.
Our strong and long background as an enterprise monitoring tool for Linux platforms ensures performance and availability monitoring of the whole end-to-end user experience for the Linux digital workspace user. Information on eG Enterprise and Linux Server monitoring can be found at Linux Server Monitoring: Red Hat, SuSE, Ubuntu, Debian (eginnovations.com).
Popular Linux distributions our customers use include Ubuntu, CentOS, RedHat RHEL and SuSE. Customers should always take care to validate that their choice of Linux distribution and version is officially supported by the GPU vendor within their guest OS support matrix, particularly with respect to GPU virtualization and sharing technologies.
GPU Monitoring for Nutanix AHV
On Nutanix AHV platforms, GPUs can also be monitored on both the hypervisor and within the VM via the “Inside VM” view. Whilst hypervisor-level view of monitoring highlights % of GPU used, VM-level monitoring highlights which application is using GPU.
The Benefits of Choosing eG Enterprise for GPU Observability
By choosing eG Enterprise for GPU enabled environments you get benefits such as:
- Monitor utilization of all popular GPU technologies from one console – NVIDIA, AMD, Intel
- Monitor across Windows and Linux platforms
- Monitor at the hypervisor level, desktop level, or at the VM level
- Identify which users are launching graphically intensive applications and their true GPU requirements
- Monitor on physical machines and cloud environments
- Make direct like-for-like assessments of GPU enabled cloud instances or on-premises VMs to optimize GPU usage and reduce costs
- Enhancements applicable for all popular digital workspace technologies including VMware Horizon, Citrix CVAD, Microsoft RDP, AVD and Windows 365 Cloud PCs, Nutanix AHV, Amazon WokSpaces and so on.
- AIOps driven root-cause analysis and automated diagnostics to help identify GPU bottlenecks
- Determine if GPU capacity is oversized to help you save cloud costs, pick the right instance type
- Those looking to leverage eG Enterprise to evaluate and compare NVIDIA vs. AMD may like to check out this good article on the subject by Marius Sandbu, see: Comparing AMD vs. Nvidia for virtual desktop GPU cards | TechTarget
- For information on what an AIOps platform offers beyond a simple monitoring tool in terms of automation, dynamic and automated metric thresholds, alerting and anomaly detection, please see: AIOps Tools – 8 Proactive Monitoring Tips | eG Innovations and AIOps Solutions and Strategies for IT Management | eG Innovations
- GPU users are frequently very demanding users of remoting protocols, as such we have recently extended eG Enterprise to support Amazon’s high-end NICE DCV / AppStream 2.0 technologies (see: Performance Monitoring for AWS NICE DCV VDI Protocol (eginnovations.com)) and provide comprehensive support for all major protocols including Microsoft RDP / RemoteFX, VMware Blast Extreme, Teradici PCoIP and CAS and Citrix HDX / ICA / EDT
- Although old, this blog is still relevant and covers the basics of NVIDIA on-prem GPU technologies for VDI and is a useful primer for those new to virtualized GPU (vGPU) technologies, see: GPU Monitoring – vGPU Monitoring | eG Innovations
- Users looking to explore AMD enabled G4ad instances on Amazon AWS may find this quick-start guide from AMD Community useful, see: The ultimate guide: setting up Amazon EC2 G4ad ins… – AMD Community
- For Citrix Linux Workspace users, you may be interested in recent monitoring enhancements added to eG Enterprise v7.2, please see: Monitoring Citrix Linux VDI (eginnovations.com)