Your passport to AWS re:Invent 2016
Reporting and analysis from IT events
Technology is progressing in the fields of augmented and virtual reality. But, for all the strengths of the cloud -- including AWS' future capabilities with high-performance computing -- it has some potential deficiencies that could prevent it from being a major player in augmented reality and virtual reality.
Virtual reality (VR) is limited mostly by the amount of processing required and the cost. VR labs must have state-of-the art equipment for high-performance computing (HPC), which can cost millions of dollars. That's well out of reach for most organizations and companies, unless they are directly involved in research or if there is a clear business case for VR. Cloud computing -- more specifically, AWS -- offers the possibility that VR or augmented reality (AR) could soon be affordable.
Now, the bad news.
Cloud computing is computing on demand, so many VR insiders don't think that AWS' future resources will contain enough processing power to drive most VR and AR applications. The popular CV1 VR interface, for instance, requires 90 Hz at 11ms per frame. If a developer thinks about sending this to AWS to update a display, then render and return the image, network latency would be a problem. Due to the limitations of the architecture, using VR may not be practical in AWS' future, regardless of the cloud provider's speed.
However, AWS does not accept this situation. Last year, Amazon brought Lumberyard onto the market. Lumberyard is a cross-platform 3D game system that allows developers to create and market high-resolution games. Thinking through this a bit, if Lumberyard proves to be useful in building 3D games that run on AWS, then VR and AR won't be much of a stretch.
Other clues exist as well. For example, I saw a job posting for a senior software developer with VR experience at Amazon Video -- not AWS. So, why is Amazon Video hiring VR talent? Well, Netflix and Hulu have VR offerings, so Amazon Video may just be trying to keep up with the Joneses. And this could lead to some direct crossover to AWS.
In addition, a Washington Post article that cites NASA's use of the Amazon cloud explains the agency faced a 100-day wait and $200,000 in expenses when using existing equipment in its on-premises data center. However, AWS was able to get to the same solution in six days for about $7,000. No additional details were provided, such as AWS' direct effect on the core VR project.
Here's the bottom line: There are some major technical obstacles to make VR and AR a reality as a cloud application in AWS' future. That said, the appearance of Amazon support for 3D gaming means these technology problems are likely solvable, and AWS' HPC cloud services could be the answer.
The use cases are many -- training police and firefighters could use VR and AR simulations; telemedicine applications are also a possibility as are 3D games and even 3D entertainment.
The question is cost. Will the price of on-premises HPC come down enough to dissuade use of AWS HPC? High-speed real-time video rendering systems are already appearing as cloud services. It's very possible that VR will be the next big application for the public cloud.
Expert advice on running HPC in the cloud
Demand for HPC in the cloud keeps growing
VR devices are the next generation of computing
The birth of the AR/VR enterprise