Virtual reality cloud is the use of a cloud service provider's infrastructure to render virtual reality (VR) code on a remote server before streaming the programming to an end user's virtual reality headset. The goal of VR cloud is to lower the total cost of experiencing VR programming by removing requirements for powerful graphical hardware and compute resources to be installed locally.
Typically, the cost of a virtual reality headset is only a third of the total cost of a high-end VR experience. In addition to the special headset that many people associate with VR experiences, there are also significant processing requirements to render 3D images and make them viewable on the headset.
Long associated with hardcore gamers and specialized training for vertical industries such as healthcare, VR and augmented reality (AR) are expected to play increasingly significant roles in industries ranging from real estate and retail to manufacturing. Goldman Sachs estimates the combined AR/VR market will be an $80 billion market by 2025, which is roughly the same size as the PC market at the time of this writing.
High end graphic processing units (GPUs) can cost upwards of $1000, and while the top end is not required, the best experience is usually delivered by distinctly high-end cards. Computer processing unit (CPU) requirements for VR programming can also be high, but rendering the scene on a remote server and then streaming it to a headset eliminates the need for the expense of a powerful CPU.
Latency and the VR cloud
While the ability to deliver VR to those who would like to experience it without significant investment may sound ideal, the caveat of VR cloud implementations is latency. Even on a local machine, gamers have been known to invest in low latency mice, keyboards and monitors. This is because even small differences in latency can give one system a competitive edge or hold a player back from playing their best.
As next-generation 5G technology rolls out, developers hope that edge gateways will be able to stream VR programming at 90 frames per second, which is the benchmark for stable VR. When throughput speeds are lower, latency can quickly cause VR sickness, a condition that leaves the end user with a feeling of disorientation and nausea.