Rob Csongor, VP and General Manager of NVIDIA's automotive business, provides his testimony on the important subject of self-driving vehicle technology.
Building a Future Where Everyone Can Ride and Drive Electric by Bridget Gilmore
NVIDIA Testimony at Senate Commerce, Science, and Transportation Committee Hearing
1. NVIDIA Testimony
June 14, 2017
Senate Commerce,Science,and Transportation CommitteeHearing
Thank you, Chairman Thune, Senator Nelson and distinguished members of the
Committee.
My name is Rob Csongor. I am vice president and general manager of NVIDIA’s
Automotive business. NVIDIA is one of the world’s leading computer technology
companies. We’re headquartered in Silicon Valley, with more than 10,000 employees
worldwide.
I appreciate your invitation to give testimony today on the important subject of self-
driving. In particular, I am grateful for the opportunity to introduce you to the
breakthrough work NVIDIA is doing in artificial intelligence. Along with hundreds of our
partners, we believe AI is the new computing model, the game-changer that makes
autonomous vehicles possible. By understanding how AI works, we can achieve better
regulatory decisions and accelerate our progress to what we all want – deployment of
safer, self-driving vehicles that will save lives.
NVIDIA’s computer innovation is focused at the intersection of visual processing, high
performance computing, and artificial intelligence — a unique combination at the heart
of the world’s next-generation computer systems.
This new form of computing is based on our invention of the GPU or graphics
processing unit, nearly two decades ago. The GPU was originally designed to power
computer graphics, but it has evolved into a powerful computer brain that processes
massive amounts of data at extraordinary speed.
Ten years ago, researchers began to use GPUs to accelerate mathematically intense
applications, such as mapping the human genome and predicting weather. More
recently, scientists working in a new field of AI called deep learning, discovered that
GPUs are critical to creating algorithms that enable computers to learn from experience
and data, similar to how the human brain works. In a short period of time, AI algorithms
rapidly outperformed code written manually by programmers. As a result, deep learning
has become a strategic imperative across many industries. Consumer services from
2. companies like Google, Amazon, Microsoft and Facebook powered by our technology
are now available to millions. In the healthcare industry AI is accelerating the search for
cancer cures. For scientists and researchers, NVIDIA delivers supercomputing solutions
used at the Department of Energy, the Department of Defense, the National Institutes of
Health, among other organizations.
The automotive world is next.
A self-driving car is an immense computational challenge. The car must be able to
detect and perceive objects everywhere around it, in motion, and in diverse weather and
lighting conditions. The car must determine its precise position, plan safe paths from
one point to another, and then drive while navigating complex situations. We simply
cannot get there with conventional programming science. AI technology can solve these
problems.
To this end, NVIDIA has created an open computing platform comprised of powerful
processors optimized for AI, in both the car and the data center. In addition, NVIDIA is
developing a full, open software stack that the automotive ecosystem is building on.
Today, we are working with virtually every automaker on research and development of
advanced self-driving vehicles using AI. Our technology is being used by more than 225
automotive companies worldwide, including Audi, Tesla, Toyota, Volvo, Mercedes and
others.
We are now at the point where we can create AI systems that have levels of perception
and performance far beyond humans, and importantly, do not get distracted, fatigued or
impaired.
Much like humans gain knowledge through experience, AI systems improve over time
with additional training data and testing.
An AI system works by training a deep neural network with large amounts of data in a
data center, monitoring and testing accuracy. Once validated, the car is updated with
new algorithms over the air, like any modern computer or mobile device. The car runs
the algorithms on real road conditions. The results are then sent back to the data
center, where the new data can be used to retrain and improve the algorithms. And then
the cycle continues, making the entire fleet better with each iteration.
Our methodology, along with our partners, will combine multiple layers of testing – in a
data center, on proving grounds, and on public roads. In addition, leveraging NVIDIA’s
3. expertise in visual computing, we can use computer simulation to accelerate the training
and testing process.
We believe new regulations are necessary. But clearly, there are opportunities to
streamline development and testing. Ideally, we would be able to test cars and collect
diverse data from any state. The patchwork of different regulations across regions
hampers that. It would be enormously beneficial to have a unified set of regulations
across all states.
It would also be constructive to ensure the standards for compliance are set correctly.
The bar we are comparing against is a human driver. A system that is significantly safer
than a human driver can save lives, once deployed. Conversely, unrealistic compliance
targets runs the risk of costing lives.
And finally, the deployment of a fleet on real roads, collecting lots of data, is the path to
achieving safety for the entire fleet.
Self-driving holds the promise to change our lives. Through our inventions, our
research, and the incredible work of development partners innovating on our
technology, NVIDIA believes this promise is achievable. We look forward to working
with this Committee, the Department of Transportation, NHTSA, and other groups to
ensure the safe deployment of autonomous vehicles through game-changing
technology paired with effective policy and regulation.
Thank you for the opportunity to tell you of our work.