When Microsoft releases its second Holographic Processing Unit or HPU, which presumably will arrive in HoloLens 2.0, it will have a custom silicon AI coprocessor for superior mixed reality.
These days, our actual reality is no longer enough. We’ve been going virtual for decades, and now augmented reality has come to life, so much so that we now talk about mixed realities.
Technological evolution is a reality of life, and so a sneak peek into the future of the HPU or holographic processing unit used in Microsoft’s HoloLens wireless headset has been granted to us all by Marc Pollefeys, Microsoft’s Director of Science for HoloLens, in a new blog post entitled “Second version of HoloLens HPU will incorporate AI coprocessor for implementing DNNs.”
A DNN is a “deep neutral network”, which can be plural as in DNNs, with Pollefeys noting that it isn’t “an exaggeration to say that deep learning has taken the world of computer vision, and many other recognition tasks, by storm,” with “many of the most difficult recognition problems have seen gains over the past few years that are astonishing.”
{loadposition alex08}
Pollefeys notes that while we’ve “seen large improvements in the accuracy of recognition as a result of Deep Neural Networks (DNNs), deep learning approaches have two well-known challenges: they require large amounts of labelled data for training, and they require a type of compute that is not amenable to current general purpose processor/memory architectures.”
He notes that “some companies have responded with architectures designed to address the particular type of massively parallel compute required for DNNs, including our own use of FPGAs, for example, but to date these approaches have primarily enhanced existing cloud computing fabrics.”
Pollefeys then states that he works “on HoloLens, and in HoloLens, we’re in the business of making untethered mixed reality devices. We put the battery on your head, in addition to the compute, the sensors, and the display. Any compute we want to run locally for low-latency, which you need for things like hand-tracking, has to run off the same battery that powers everything else. So what do you do?”
The answer is that “You create custom silicon to do it,” with this video then displayed.
Pollefeys goes on to give additional background on the HPU, with all the tech and sensors that go in to making the HPU, which inside HoloLens makes it “the world’s first–and still only–fully self-contained holographic computer.”
We then get some extra detail on how Microsoft’s Harry Shum, the executive VP of the company’s “Artificial Intelligence and Research Group, announced in a keynote speech at CVPR 2017, that the second version of the HPU, currently under development, will incorporate an AI coprocessor to natively and flexibly implement DNNs.”
On display was an “an early spin of the second version of the HPU running live code implementing hand segmentation” with a range of layer types that are fully programmable by Microsoft’s HoloLens creators.
Pollefeys concludes by stating that “The AI coprocessor is designed to work in the next version of HoloLens, running continuously, off the HoloLens battery.
“This is just one example of the new capabilities we are developing for HoloLens, and is the kind of thing you can do when you have the willingness and capacity to invest for the long term, as Microsoft has done throughout its history.
“And this is the kind of thinking you need if you’re going to develop mixed reality devices that are themselves intelligent. Mixed reality and artificial intelligence represent the future of computing, and we’re excited to be advancing this frontier,” he concludes.
So, although a successor version is obviously no secret in being under development, we’ve been given a sneak peek into the advancements a second-gem HPU will deliver - and it is no doubt keeping Apple, Google, Facebook and Oculus, HTC, Samsung, Magic Leap and everyone else on their toes with developments of their own yet to come!