Nvidia tried to find its luck with music and video players, later with mobile phones and it abandoned both of the markets as the market aromatized much faster than most expected. A few years back Nvidia started putting Tegra SoCs in cars and that strategy worked out really well.
Jen-Hsun Huang, President and Chief Executive officer at Nvidia, has confirmed to analysts that Nvidia shipped in 5 to 6 million cars (probably in the last quarter). In August 2015, Nvidia confirmed that the company had a 70 percent year over year growth and that it sold SoCs that are shipping in 30 million cars. The growth number got slightly adjusted two quarters later as Nvidia how claims 68 percent growth YoY.
The company also claims that they have additional 20 to 25 million units to ship in its pipeline. These are design wins that took quite few years of engineering to ramp into production. This gives Nvidia good visibility of the pipeline and the opportunities that are ahead. In other words, the future is bright. Computerized car became a highly desirable end user feature and Audi and Tesla owners have proven that. Honda, for example, in Europe offers a 7-inch Android powered navigation for additional €700 which a small price to pay to get rather functional infotainment system with navigation. These things just a few years back use to cost between €2,000 to €3,000 and it was not touch enabled or as powerful as the infotainment of today.
The second part of the Tegra car revenue story is yet to happen. Nvidia has a great chance to earn a lot of money from the self-driving the next step in development of ADAS (Advanced Drive Assistance System) strategy. Nvidia's solution is called DRIVE PX and DRIVE PX 2. This is a rather complex subject and Nvidia claims that it works with quite a large number of customers now, car companies, start-up companies, companies that are largely cloud-based and have an enormous amount of data that they could transform into an automotive service, transportation as a service.
Nvidia benefits a lot now that self-driving vehicles require much more computational power than the ADAS, and in each one of those levels of autonomy, a different amount of computation would have to be deployed. Nvidia has created a scalable architecture that allows car companies to develop cars that are partially assisted, all the way to completely assisted.
When self-driving finally reaches the consumer affordable vehicles Nvidia should be able to sell three different products inside of the same car. One SoC for the dashboard computer, one SoC solution for the infotainment – navigation and one SoC based DRIVE PX for the self-driving feature.
Of course, Qualcomm will try to take some of this market and the company is launching its first infotainment and dashboard solution with a few customers but it won't be easy for Qualcomm to attack the self-driving market. We see a bright future for Nvidia Automotive, this part of the company really has a lot of space to grow and can easily even multiply what the company is making right now in this market.
