Advertisement
Advertisement
Lenovo products on display during the company's recent annual results announcement. Photo: Nora Tam

Machine vision provider Movidius teams up with Lenovo for virtual reality products

Visual processor provider Movidius is partnering with computer giant Lenovo to develop virtual reality (VR) devices, the company said Tuesday.

The San Mateo, California-based company provides chips that help computers process more complicated visual information, including comprehension of gestures, recognising faces and sensing space and depth.

Visual recognition technology is on the verge of spreading widely to many industries such as consumer devices, security cameras and autonomous driving, according to Lin Renxiang, an analyst at Chinese research firm HCR.

“We want to be to [artificial intelligence] what Intel is to the PC and what Qualcomm is to mobile,” said Movidius chief executive Remi El-Ouazzane at the RISE conference in Hong Kong. Intel and Qualcomm are leading processor providers for desktop computers and smartphones, respectively

Under the partnership, Movidius provided its Myriad 2 Vision Processing Unit (VPU), a dedicated chip that processes visual information, to Lenovo which incorporated the technology into a VR product line it will announce at the Lenovo Tech World event in San Francisco on June 9.

El-Ouazzane declined to reveal the new Lenovo products but said the collaboration will result in several VR products.

According to Movidius, its technology can be used in VR devices including head-mounted displays, multi-lens cameras and hand-held controllers.

Movidius chips help enable products like virtual reality head-mounted displays. Photo: Xinhua
The VPU technology is similar to how the current personal computer is configured. A central processing unit, or CPU, is responsible for most of the general tasks while other processing duties are handled by separate chips, like the graphics processing unit or GPU that handles all the complex graphics tasks such as video games and video editing.

Movidius said its VPU will offload vision related tasks from the CPU and GPU to itself and improve the overall performance of a given system.

This is not the first deal Movidius has struck with a Chinese technology heavyweight. It also provided the VPU to leading drone maker DJI for its latest Phantom 4. Released in March, the drone can track users on the move while automatically avoiding obstacles along the way.

Lin from HCR believes machine vision will become the backbone of the smart cities of the future where visual identification, traffic monitoring and self-driving cars all require this technology.

Movidius is not alone in this field of vision recognition. Technology companies such as Nasdaq-listed Synopsys and NYSE-listed Mobileye also offer their own machine vision solutions.

El-Ouazzane said the company will soon announce a partnership with a major Chinese provider of security cameras and is on its way to become a listed company in two years.

Movidius completed its series E funding of US$40 million last year with investors including Chinese optical manufacturer Sunny Optical Technology. It has a research centre in Shanghai and an office in Hong Kong.

Post