NVIDIA began working with automotive customers well over 10 years ago. Today, the overall NVIDIA strategy in automotive is focused on self-driving cars.
At the time that the company began a dedicated business in automotive, the focus was placed on in-car entertainment and display systems. The relationships with automotive OEMs was a natural extension to the relationships the company had already developed through the professional workstation business with NVIDIA Quadro GPUs.
Today, automotive OEMs are examining many different solutions. They are using super-computing in the data center and NVIDIA enables this computing. They are using platforms like NVIDIA Drive PX, the Xavier SOC for in-car processing and artificial intelligence. The new generation of Xavier was launched at CES 2018. The platform is a highly complex SOC product that can go in Level 3, level 4, and level 5 self-driving cars. The performance and energy requirements provide the necessary levels of performance and power. It is about 30W and 10x better than the previous generation. The platform is possible to be used in prototypes, but also available to be used in production vehicles.
Today, the company has more than 320 partnerships in the automotive market. These partnerships are important to help the company understand the dynamics in the industry, globally, regionally, and with specific challenges in self-driving cars. Self-driving cars used to be a 2030 target for automotive OEMs, and now the targets have moved closer to 2020.
A single product turns into multiple solutions
This strategy in developing specialized market solutions and expanding the addressable TAM is not a new strategy for the company.
NVIDIA has one "product" which is the GPU. That same product that delivers great gaming experiences, professional visualization, and computing in the data center delivers performance for a very wide range of application areas. The strategy of NVIDIA is to make the huge investments needed in a powerful GPU architecture and leverage that investment across multiple, high-performance accelerated-computing markets. The single architecture is leveraged with a single development platform which includes the NVIDIA programming environment, CUDA. Then each of the specific markets is addressed with additional software tools, SDKs, and with specialized hardware solutions, such as Xavier and NVIDIA Drive. The huge number of partnerships in automotive is also an old habit of the company. NVIDIA has always worked with a range of partners in every market they support. They have the most extensive relationships with software vendors (ISVs) in the professional workstation markets and this has driven the Quadro product to maintain an 80%++ market share for over 15 years.
The natural approach to developing this rapidly moving autonomous vehicle market is to seed the market as widely as possible. The company clearly believes that its solution is better adapted than other options to solving the hard problem of autonomous driving. They view the autonomous vehicle market as a huge market. They are probably correct about the former, and they are clearly correct about the latter.
The PW Perspective
The company is in front of the competition in providing a viable solution. They are building a wide network of partners. They are engaging with every significant company. And they deliver not only the in-car solutions, but they deliver the full pipeline of AI-enabled computing from data center solutions to desk-side solutions to in-vehicle solutions.
