PWW

Portfolio AI Insights

pww.comNVIDIA Corporation Presents at Needham & Company's 4th Annual Automotive Tech Conference, Jun-03-2020 11:15 AM - NasdaqGS:NVDA

NasdaqGS:NVDA

Danny Shapiro [Former VP of Automative] 💬

During the presentation, Danny Shapiro, the Senior Director of the Automotive Division at NVIDIA, discussed the following points:

  1. Introduction to NVIDIA's Work in the Automotive Industry:

    • NVIDIA started as a gaming company but has evolved to be at the forefront of the AI revolution.
    • There are two aspects of AI: training and edge AI/inference.
    • NVIDIA's DGX systems are used for training AI in data centers, and the same architecture is used in vehicles through the NVIDIA DRIVE AGX platform.

  2. Overview of NVIDIA's Solutions for Autonomous Vehicles:

    • NVIDIA provides an end-to-end platform that includes data collection, training, and testing/validation.
    • The platform supports varying levels of automation and autonomy across different types of vehicles.
    • NVIDIA's solutions include software stacks like DRIVE AV for autonomous driving and DRIVE IX for intelligent experiences within the vehicle.

  3. Deep Neural Networks in Autonomous Driving:

    • NVIDIA develops dozens of deep neural networks for tasks like object detection, free space understanding, and weather condition recognition.
    • These networks are computationally intensive and require significant amounts of data for training and refinement.

  4. Simulation and Validation:

    • NVIDIA's DRIVE Constellation platform allows for hardware-in-the-loop and software-in-the-loop testing and validation.
    • The platform uses virtual reality to generate synthetic sensor data and simulate environments for testing hazardous scenarios.

  5. Software Stack and Hardware:

    • NVIDIA provides a comprehensive software stack that includes DRIVE OS, DRIVE AV, and DRIVE IX, along with tools, libraries, and deep neural networks.
    • The hardware includes the DRIVE platform, which is used by hundreds of companies and is scalable and backwards-compatible.

  6. New NVIDIA Orin Chip:

    • NVIDIA Orin is a 17-billion-transistor chip that offers a 7x performance boost over the previous generation.
    • It combines different processors, including ARM CPUs, Ampere GPUs, and deep-learning accelerators.
    • Orin can support solutions ranging from entry-level ADAS to high-end robotaxis.

  7. Unified Architecture:

    • NVIDIA's unified architecture allows automakers to implement a single solution across their entire vehicle lineup.
    • This approach simplifies software development and enables over-the-air updates to add new features and capabilities.

  8. Testing and Validation in Simulation:

    • Due to the COVID-19 situation, NVIDIA shifted to simulation-based testing, allowing engineers to continue working remotely.

  9. Ecosystem:

    • NVIDIA works with hundreds of companies, including OEMs, startups, and software companies, to develop and deploy autonomous vehicles.
    • The ecosystem includes partnerships with Tier 1 suppliers, mapping companies, and sensor companies.

  10. Differentiators of NVIDIA's Platform:

    • NVIDIA's end-to-end platform covers all aspects of autonomous driving, from data center to vehicle.
    • The single architecture spans from ADAS to robotaxis, reducing the need for porting code.
    • The open software platform allows developers to customize and build their own applications.

  11. Market Engagement:

    • NVIDIA engages with customers at different levels of automation, from Level 2+ to Level 4 and Level 5 robotaxis.
    • The market for Level 2 and Level 3 vehicles is expected to grow significantly over the next decade.

  12. Technology Discussion:

    • The migration to NVIDIA's Ampere architecture offers 4x the performance and power efficiency over previous solutions.
    • The performance gains come from a combination of process node improvements and optimized algorithms.

  13. Robotics and Trucking:

    • Robotics represents a different segment from automotive but holds significant growth potential.
    • NVIDIA sees a strong market opportunity in automating logistics and factory operations.

  14. Product Portfolio Feedback:

    • The feedback from OEMs regarding NVIDIA's product expansion is positive, especially regarding the single architecture.
    • Car manufacturers appreciate the ease of use and the ability to update software over the lifecycle of the vehicle.

  15. Conclusion:

    • NVIDIA's ability to train AI in the data center and directly apply it to vehicles for inference is a significant differentiator.

Kimberly Powell [Vice President of Healthcare] 💬

** Key Points Mentioned by Kimberly Powell**

  • Market Strategy for Higher-End Systems: Kimberly Powell discussed the deployment of higher-end systems in more vehicles, even if the software isn't initially activated. She highlighted a company that has adopted this strategy, placing an AI computer and sensors in cars to prepare for future Level 4 capabilities.

  • Potential for Future Upgrades: She emphasized that car companies need to adopt this strategy, where vehicles are equipped with the potential to be upgraded to Level 4 autonomy as soon as the software and regulations are in place.

  • Consumer Demand for Software-Defined Vehicles: Kimberly noted that once consumers start to see the benefits of software-defined vehicles, they will prefer them over fixed-function devices, similar to how they prefer smartphones that receive regular updates.

Feedback