Enterprise Virtual Reality (VR) and Augmented Reality (AR) just took another leap forward.

TL;DR: With the launch of NVIDIA CloudXR SDK 1.0, VMware is partnering with NVIDIA to securely deliver AR and VR applications running on VMware vSphere, and streamed using NVIDIA to VMware’s Project VXR client application running on standalone headsets. Project VXR is an xLabs innovation project within VMware’s Office of the CTO. We recently published a whitepaper outlining our results on “Running AR/VR Applications on VMware vSphere with NVIDIA CloudXR”.

Enterprise AR/VR, aka spatial computing, has seen growing traction within the enterprise over the last 18 months. The growth has been driven by a combination of an increasing number of available devices, improvements in hardware and a maturing ecosystem. Enterprises have woken up to the fact that spatial computing delivers a number of key productivity gains, particularly in a time of increased remote working.

Benefits of Spatial Computing in the enterprise

However, delivering high-fidelity graphics and high framerates for AR and VR has been limited to physical workstations where the headsets were tethered. Recent standalone headsets made VR more available to consumers and enterprise users, but with the limitation of low power CPUs and GPUs, this means AR and VR experiences lacked fidelity.

The Need of High-Fidelity VR Experience

Training is the top VR use case in the enterprise and for many customers, high-fidelity simulations are critical, and many customers have already invested in PC based VR experiences. Design visualization is another key use case in the enterprise for VR and again high fidelity is a critical aspect to this.

 

Companies such as RIVR offer VR training for fire investigation, requiring high fidelity simulation

Using mobile devices such as phones, tablets and even Microsoft’s HoloLens 2 headset for AR also suffer similar limitations to standalone headsets. Viewing complex high-fidelity models in AR is critical in use cases such as design reviews, inspection, repair and assembly, and standalone devices struggle.

Enterprise AR/VR with VMware and NVIDIA CloudXR

Now with VMware and NVIDIA CloudXR, enterprises can get the best of both worlds, the high-fidelity graphics and framerates of powerful PCs and with the ease of use and mobility of standalone headsets and devices.

 

Remote AR/VR solution using VMware Project VXR, NVIDIA CloudXR and VMware Horizon with NVIDIA GPUs

We are proud to be partnering with NVIDIA to securely deliver AR and VR applications running on VMware vSphere, streamed using NVIDIA CloudXR to VMware’s Project VXR client application running on standalone headsets.

Streaming AR/VR over Wi-Fi or 5G enables organizations to truly take advantage of the benefits AR and VR offer. Instead of having dedicated physical VR workstations, organizations can simply provide low cost standalone headsets to users. AR/VR applications run on VMware vSphere in the datacenter or at the edge and streamed to users for easy access. There are many other benefits to centralizing your AR/VR applications such as security, manageability, secure access for partners and suppliers, no need to distribute large datasets and the portability of virtualization (i.e. push workloads to other locations).

Note: In the diagram above VMware Horizon is used for desktop management only and does not include CloudXR functionality today.

Is this just hype?

In a survey of over 700 enterprises by Cap Gemini, 600 (85%) were actively deploying or evaluating AR or VR. At VMware we have our own internal deployments for AR and VR and we’re seeing an increasing number of our customers needing our help in deploying and managing AR/VR infrastructure.

Despite growing traction in the enterprise, adopting spatial computing is challenging. Devices today have come from the consumer space and lack enterprise features. Hardware vendors such as Facebook/Oculus, Pico and HTC are moving towards enterprise support, but ultimately these devices need to integrate with existing IT systems, and that’s where VMware and specifically Workspace ONE and VMware’s Project VXR come in.

Enterprise AR/VR Solution

VMware can provide an end-to-end solution to address current deployment challenges, from device management and secure application access, to remote rendering and infrastructure management.

Any enterprise AR/VR solution needs the following key components (we’ve included the products we are using):

  • Device Management – VMware Workspace ONE
  • Enterprise AR/VR Client – VMware Project VXR
  • Secure Application and Content Distribution and Access – VMware Project VXR and VMware Workspace ONE
  • Remote AR/VR capability – VMware Horizon on VMware vSphere with NVIDIA CloudXR and Project VXR
  • Single Sign-On – VMware Project VXR and VMware Workspace ONE

These pillars represent key functionality needed for any enterprise deployment and will be explored in a future blog on enterprise AR/VR architecture.

Does AR/VR Streaming Really Work?

We’re all familiar with the challenges in streaming video let alone high-fidelity interactive experiences. AR and VR have strict latency requirements in order to ensure the user is comfortable during the experience. The Project VXR team worked closely with our colleagues across VMware, including VMware’s  HCI Chief Technologist Duncan Epping. Duncan helped with validating our enterprise AR/VR solution, including running AR/VR applications on VMware vSphere and streaming using NVIDIA CloudXR to Project VXR’s client running on a standalone headset.

VR design review experience at 72FPS on Oculus Quest with a model with 17,000,000 polygons

In our white paper outlining this test, we validated using VMware vSphere® backed by NVIDIA Virtual GPU technology. We showed that AR/VR applications can run on a Windows 10 virtual machine with an NVIDIA vGPU, and streamed to a standalone AR/VR device, such as the Oculus Quest or Vive Focus Plus, using NVIDIA CloudXR’s protocol.

Our testing validated that streaming over 5Ghz Wi-Fi provided an excellent user experience at maximum frame rates on devices such as the Oculus Quest that run at 72 Frames Per Second (FPS). NVIDIA CloudXR handled latency spike over the Wi-Fi connection to prevent any impact on the user experience. When running two VR applications on a single GPU we maintained 75FPS on each virtual machine. The only complication we ran into was the number of encoders on the GPU limited further number of sessions running on a single GPU.

 

Performance results from running AR/VR applications on VMware vSphere and NVIDIA GPUs using vGPU technology.

As more organizations explore AR and VR solutions for training and improving efficiency in their workplace, Remote AR/VR will become a critical technology. Remote AR/VR will be crucial to improve the mobile AR/VR user experience, and the manageability and security of these workloads.

If you are currently exploring an enterprise AR or VR deployment, we would love to collaborate with you. We invite you to connect with us here (via the pop up window).