Home Interviews NVIDIA is leading the world in self-driving technology

NVIDIA is leading the world in self-driving technology

NVIDIA is leading the world in self-driving technology

NVIDIA has created a leadership position for itself in developing technology to power self-driving cars. Vishal Dhupar, Managing Director – South Asia, NVIDIA Graphics Pvt Ltd, explains the company’s capabilities in the automobile industry and beyond. Excerpts:

What is range of products offered by NVIDIA in India? What are the key features and capabilities of your products for small, medium, large and luxury cars?

There are several facets to NVIDIA’s automotive business, which fall broadly into three categories: High-end graphics used in product design, infotainment and autonomous driving. On the graphics side, we work with a wide range of automotive manufacturers and Tier-1 suppliers, from Jaguar Land Rover, Porsche and Audi to Ford, Honda and Skoda.

Every automaker in the world uses NVIDIA’s Graphics Processing Unit (GPU) technology in the design and styling of their vehicles, from the computer aided design (CAD) models employed by engineers to virtual prototyping. GPUs are also widely used in the marketing of vehicles, most spectacularly in the impressive Audi City showrooms, where consumers can visualise their new car life-size with any combination of features, accessories and colours on NVIDIA-driven power walls.

Inside the car, NVIDIA GPUs provide a beautiful and intuitive experience on every screen, including virtual instrument clusters, navigation and entertainment systems. Examples unveiled at the recent Geneva Motor Show include Audi’s new Q2. This SUV is aimed at a new generation of digital natives. Options include the all-digital Audi virtual cockpit and a head-up display and a stunning NVIDIA-powered navigation system with online Audi Connect.

Porsche’s 718 Boxster features an all-new touchscreen infotainment and navigation system powered by NVIDIA. And BMW’s latest M760Li includes plenty of power to go with its advanced NVIDIA-powered cockpit.

Over the last two years, NVIDIA has also created a very niche position for itself in developing technology to power self-driving cars. Its DRIVE PX platform is a supercomputer, the size of a lunchbox, that delivers the massive computing power required by the artificial intelligence systems which will navigate our cars in the future. Audi is the first automaker to announce its developing advanced driver assistance functionality based on DRIVE PX, which will appear in the Audi A8 next year.

How do you differentiate your product features?

NVIDIA’s unparalleled graphics heritage, parallel processing expertise and 20-year track record of working with automakers has given us a unique leadership position in the market. Consumers now expect the same level of interactivity, functionality and beauty from the technology in their car as they do from their smartphone or tablet, so it’s natural for OEMs to turn to the world leader in visual computing to help them achieve this.

When it comes to self-driving technology, no other company can do what NVIDIA does. Over the lifetime of our company, we have invested billions of dollars worth of R&D in our core technology, the Graphics Processing Unit. This same tiny piece of silicon that creates stunning graphics for gaming, professional graphics applications and in-car infotainment systems is also highly suited to the computing required for deep learning, the branch of artificial intelligence that will soon make self-driving cars a reality. DRIVE PX 2, which packs processing power equivalent to 150 Macbook Pros, will be the brain of your self-driving car.

Do your products require connectivity with IOT, DSRC and such other networks for optimum performance? What are the areas of infrastructure facilities development one would require in future, to fuel the acceptance level of infotainment products by
the users?

While connectivity will be an important factor in making self-driving cars a reality, it’s not feasible for an autonomous vehicle to be reliant on constant connectivity. A car which loses control if it hits a patch of bad cell signal or goes through a tunnel is not acceptable. That’s why we believe the computing power required for self-driving functionality must be on board, not just in the cloud.

One side effect of self-driving cars will be the freeing up of in-vehicle screen real estate and the ability of the car’s occupants to consume infotainment content. Concept vehicles, such as the Mercedes F 015 Luxury in Motion imagine the car as a ‘mobile living space’ where every surface is a screen capable of displaying rich interactive content.

What are the challenges related to the further development of your products? What do they entail in terms of new technologies, design concepts, big data management and security of your systems from hackers?

NVIDIA is leading the world in self-driving technology and we’re currently working with more than 70 companies and research institutions on autonomous vehicle projects. Although automotive is the smallest division in our company, it’s also the fastest growing. We are able to meet the challenge of resourcing this opportunity because we can leverage the expertise of employees from across multiple divisions in NVIDIA.

Security is certainly a major topic which must be considered in the development of autonomous vehicles. Last year, hackers demonstrated they were able to take control of a Jeep Cherokee remotely, exploiting a security loophole in the vehicle’s wireless connection. Such issues can arise when a vehicle’s computer systems are not effectively integrated. In the past, one approach to adding technology functions to cars was simply to ‘bolt on’ new hardware or software components, creating an infrastructure that was riddled with potential entry points at risk of exploitation. For this reason, NVIDIA works with auto makers to develop centralised car computers with the power to handle future functionality and
the security framework to maintain
their integrity.

Leave a Reply

Your email address will not be published.