Amazon has long been rumored to be developing a broad portfolio of virtual reality products and related services, including VR movies, a platform and interface for content, and even an expansive virtual shopping mall. This week, the Seattle-based conglomerate announced G3 Instances, the latest generation of Amazon EC2 Accelerated Compute Instances. The announcement is important as application developers can save time and money when configuring infrastructure by now launching G3 instances, a web service that provides resizable compute capacity in the cloud, to support graphics-intensive applications. G3 instances will make it easy to procure a powerful combination of GPU, CPU, and host memory for workloads that include 3D rendering, 3D visualizations, graphics-intensive remote workstations, video encoding, and virtual reality applications.

The Amazon G3 Instances combine NVIDIA Tesla M60 GPUs and Intel high-performance processors, doubling the CPU power per GPU and host memory per GPU when compared to the most powerful GPU cloud instance available today. This allows users to perform complex modeling and 3D visualization analyses such as medical image processing, computer-aided design, or seismic visualization jobs in much less time than what was possible with other GPU cloud instances.

While it is unclear how much more G3 instances cost compared to EC2 instances, the new service changes the economics of computing and removes friction. Moreover, G3 instances can be obtained and configured using the AWS Management Console, allowing the developer to get started with minimal friction. G3 instances are currently available for servers in the U.S. and Ireland, but more regions are expected to come online soon.


Clifton Dawson, CEO of Greenlight Insights, is an analyst who covers digital disruption and how today’s company are transforming their business models to deliver next-gen experiences through virtual and augmented reality technology. Follow him: @AskClifton.