NVIDIA’s end-to-end autonomous vehicle solution
NVIDIA (NVDA) is a leader in the autonomous driving market with its advanced end-to-end solutions. Intel (INTC) is catching up in this space but is still far behind NVIDIA in providing a complete end-to-end solution. NVIDIA has expanded its autonomous driving market to include trucks, mobility services, and even construction and industrial vehicles like forklifts.
Ready to put your morning scrolling to use? Sign up for Bagels & Stox, our witty take on the top market and investment news straight to your inbox! Whether you’re a serious investor or just want to be informed, Bagels & Stox will be your favorite email.
At the 2019 Investor Day, NVIDIA’s vice president and general manager of its Automotive business, Robert Csongor, explained NVIDIA’s three-phase end-to-end AV (autonomous vehicle) solution. He explained that the first phase is a supercomputer that will go in an AV. This market presents a $25 billion TAM (total available market) by 2025, as it will include cars, trucks, robo-taxis, forklifts, and many other types of vehicles.
NVIDIA is leveraging its Xavier platform to develop various types of solutions from its Level 2+ autopilot solution called DRIVE AP2x to Level 3 and 4 AV solutions. Robert Csongor stated that the Level 2+ solution alone presents a $17 billion TAM, as it can be sold as an independent solution that drivers can attach to their cars. He gave the example of Tesla’s (TSLA) autopilot function, which has an attach rate of ~80%.
Training and development
Robert Csongor stated that the second stage is the training and development stage where the supercomputer that will go in the AV needs to be trained on DNNs (deep neural networks). NVIDIA has been collecting millions of images and labeling them as the path, pedestrians, signs, and cars to create DNNs. Robert Csongor stated that each car needs to be trained on more than ten DNNs to make it drive ready, which presents a $3 billion TAM for NVIDIA’s DGX supercomputer, a leader in deep learning.
Robert Csongor stated that the third stage is validation where the trained computer needs to be tested to ensure the car that is put on the road is safe. As validating an AV in the real-world environment will take a long time, NVIDIA devised a solution of creating a simulated environment where developers can create challenging scenarios to test the AV.
NVIDIA’s DRIVE Constellation HIL (hardware in the loop) uses two computers, one to create the simulated environment, and the second to drive the car. Its DRIVE Constellation SIL (software-in-the-loop) allows resimulation on new data. Validation presents a $2 billion TAM by 2025.
Check out all the data we have added to our quote pages. Now you can get a valuation snapshot, earnings and revenue estimates, and historical data as well as dividend info. Take a look!