With such a vast space to work in now, virtual reality, as well as augmented reality has turtled developers and headset builders, forcing them to come up with creative and genius ways to overcome some of the boundaries imposed, such as haptic devices for tracking movements, a means to travel across digital environments while maintaining ‘presence’, ways for application to meet the required 90fps to reduce the nauseous side-effects and more unforeseen dev issues are rising.
Some development studios are attacking some of these issues at the hardware level, which is exactly what Movidius is aiming to do. Movidius was started in 2006 by highly skilled, highly experienced experts in the field of computer processing units. Their intent, based off the absence of power-efficient micro-chips that could give computers( this includes PCs, smartphones & smart-devices) vision cost-effectively and Moore’s Law slowing down, was to make this a reality.
After a decade, they have serviced many clients and have developed a suitable Visual Processing Unit(VPU), with their current iteration being the Myriad 2. Myriad 2 is a System-on-Chip(SoC), that handles all the visual input, as would the Graphics Processing Unit(GPU) in your computer would handle all the graphics. Their SoC also contains their proprietary processor, SHAVE (Streaming Hybrid Architecture Vector Engine), which insures the Myriad 2 is hi-performing and power efficient.
So how does this benefit VR/AR headsets? One of the key elements of succesful ‘presence’ in mixed reality, relies in how we interact within the digital environment, which is aiming more and more towards full body recognition. In other words, it’s better that we kick a CGI ball with our own feet, rather than pressing buttons on a controller.
Truth is, these sorts of input require a lot of processing power for the computer to understand such movements, which is exactly the solution their SoC Myriad 2 provides. In their words, “Myriad 2 allows VR and AR devices to crunch huge amounts of data at low power and ultra-low latency.”
Because of their huge accomplishments in giving computers ‘vision,’ another step in fully fledge artificial intelligent software, Intel has decided to acquire them. When it comes to Intel’s initiative with Realsense, they aim to stay ahead of the curve, because “with Movidius, Intel gains low-power, high-performance SoC platform for accelerating computer vision applications.”
They understand, that SoCs will be extremely vital, especially as the growing number of connected smart devices continue to grow, with projections of 50 billion IoT devices to be connected 2020. They also understand, that it will play a crucial part in developing proficient machines, such as robots and drones, so they can navigate or recognize the environment as a human would.
With augmented reality applications, digital objects would merge and interact seamlessly with the immediate physical environment. There are of course other unforeseen solutions Movidius VPU, the Myriad 2, will provide, though for now they are gaining a significant development boost by partnering up with Intel. You can connect with Movidius via Twitter @Movidius.
Thanks for reading! Let me know your thoughts in giving IoT devices human vision with a comment below:)