Xiaomi: Why Do We Do VR? Lei Jun Warns that Prepare for at least Five Years

The Xiaomi MiLab published an article titled “Why Do We Do VR?” on its official Weibo account on November 11. The article disclosed some stories behind Xiaomi‘s VR layout.

On January 8, the eve of CES 2018, Xiaomi officially announced its cooperation with Oculus and said it would release the Chinese version of Oculus Go. The poster showed that Xiaomi VR looks like Oculus Go. Also, the core hardware of Xiaomi VR is exactly the same as the Oculus Go, including the Fast-Switch 2K screen and Qualcomm Snapdragon processor.

SEE ALSO: Working with Xiaomi: Oculus to Release VR Devices in China

Xiaomi‘s MiLab published an article titled “Why Do We Do VR?” on its official Weibo account on November 11. In the article, MiLab told some stories behind Xiaomi‘s VR rollout. At the end of 2105, Xiaomi established MiLab and started to work on VR, AI and other fields. When Lei Jun learned about it, he said, “You must be ready and prepare for the next five years.”

“Since 2015, we have made multiple visits to the Facebook Oculus team. We learn their understanding of VR. In order to bring VR to more people, we offered to cooperate with Oculus, and finally the Oculus Go/Xiaomi VR program was established,” the article said. The success may owe much to the help of Xiaomi‘s former VP, Hugo Barra. Barra joined Facebook and took charge of its VR team in January 2017.

On the current development of VR, MiLab said the VR industry is just beginning.

“We predict that, even at the hardware level, VR will go through two smartphone-level redefinitions. In other words, there are quite a number of variables. It is even true for the most basic tracking scheme. In 2015, Lighthouse seemed to be the future, but inside-out tracking has become the obvious right direction,” it said.

At the same time, the article also expounded on the current situation from three aspects: vision, movement and content. Leifeng.com selected the following excerpts:

On the visual level, MiLab hopes that eventually the world we see in VR like in reality – that is, without the limitation of field of view and definition. The basic paradigm of imaging is still screen and optical lenses. But as the complexity and the demand for light grow, the industry will pay more attention to modeling and rendering to simulate the real world when switching between augmented and virtual realities.

VR screens only recently satisfied the demand for low afterglow: now the industry must work on resolution. But at the same time, the demand for rendering power has increased exponentially. The challenge in this aspect is great. AMD has proposed a goal of 16k per eye at 144Hz, which is no different from the real world. That would be 256 times beyond modern calculation power, which is limited to 1k per eye at 90Hz. Even if Moore’s law is fulfilled every year, it would take 10 years to achieve that goal.

On the mobile side, a proper solution can handle most problems related to VR dizziness. But there is still no fully defined paradigm. It is too common to rely on virtual transfer or analog control, and the hand movements required in Climb and Lone Echo are only available in a few scenarios. For example, in terms of the virtual running machines that simulate real running, consumers will be reluctant to buy in unless the form is practical enough. The experience itself is not satisfying, and it is not clear whether developers can meet their expectations. Whether it’s a large MMORPG or a social experience, a virtual world that people remain in requires a better method of control.

In terms of content, the development of VR has also shown some interesting things in recent years. For social platforms like Facebook, social networking is an infrastructure, and even ubiquitous, so it has to be very neutral. Products like VRChat, however, may be the social apps that the industry needs. These products are novel, absurd and boring, but interesting. VRChat lets user play as Mario, Saber and other classic cartoon characters. Cartoon fans around the world can meet and party. Although the images are still rough, they are very interesting.

Such content already has an audience in the live broadcast platform, which is compatible with non-VR platforms, but it also reflects the advantages of VR. Perhaps the 90 percent of non-VR users are envious of the 10 percent with VR hardware. In the end, social networking will not be a single “app” in virtual reality, but will appear everywhere – just as in real life.

Finally, the article also mentions a concept from Sword Art Online: in the virtual world, when the information is as intense and strong as the real world, we cannot distinguish between real and virtual. To achieve this, we need a lot of technical support and will go through a lot of uncertainties. But the obsession with exploring “another world” will always be there.

This article by Tian Miao originally appeared in Leiphone and was translated by Pandaily.