April 4-7, 2016 | Silicon Valley DRIVE ™ PX 2 SELF-DRIVING CAR COMPUTER Shri Sundaram, Spring 2016
SELF-DRIVING AND AI SUPERCOMPUTING 2
SELF-DRIVING CAR PLATFORM DRIVEWORKS Perception Localization Planning Visualization NVIDIA DRIVE™ PX 2 NVIDIA DIGITS NVIDIA DRIVENET 3
NVIDIA DRIVE™ PX 2 12 CPU cores | Pascal GPU | 8 TFLOPS | 24 DL TOPS | 16nm FF | Scalable from 1-4 processors World’s First AI Supercomputer for Self -Driving Cars 4
INTERFACES 70 Gigabits per second of I/O Sensor Fusion Interfaces: GMSL Camera, CAN, GbE, BroadR-Reach, FlexRay, LIN, GPIO Displays and Cockpit Computer Interfaces HDMI, FPDLink III and GMSL Development and Debug Interfaces HDMI, GbE, 10GbE, USB3, USB 2 (UART/debug), JTAG Auto Grade connectors Debug/Lab interfaces 5
COMPUTATION ENGINES 24 DL TOPS, 8 TFLOPS, high performance CPU complex Dual CPU-GPU Cluster Connected over • Scalable + Redundancy Capable Gigabit Ethernet Each GPU has • Independent and specialized dedicated memory compute Specialized • Greatly accelerate DNN Instructions for performance Discrete GPU DRIVE™ PX 2 6
SOFTWARE A full stack of rich software components NVIDIA Vibrante Linux & Comprehensive BSP Rich Middleware SDK, Samples and more 7
DRIVE™ PX 2 Designed For SENSOR PROCESSING COMPUTATION Sensor Fusion, Point Cloud etc. Detection/Perception (DNN Inference), Localization, Path Planning, Visualization COMMUNICATION DATA COLLECTION Automobile Network Algorithm development, Hardware in Loop etc. 8
DRIVE™ PX 2 Built for DEVELOPING RAPID READYING FOR APPLICATIONS PROTOTYPING PRODUCTION Develop applications using Validate algorithms in car DRIVE PX™ 2 as reference DRIVE™ PX 2 with live cameras/sensors for custom ECU Deploy a pre-trained DNN Partition existing system Safety concept to derive on DRIVE™ PX 2 (say, PC); include DRIVE™ certain Safety Integrity PX 2 for sensor fusion Level Migrate applications from PC to DRIVE™ PX 2 10s of cars Typically in a lab set up or a few cars 9
DRIVE™ PX 2 Benefits CORE DRIVE™ PX 2 TOOLS & SPRINGBOARDS CAPABILITIES ECOSYSTEM TO PRODUCTION Automotive SW Computation & (RTOS, AUTOSAR & Memory Code Compliance) Rich SW & HW Capabilities Functional Safety Product Partnerships Interfaces & I/O Automotive HW Bandwidth (HW & SW) 10
CORE DRIVE™ PX 2 CAPABILITIES COMPUTATION & MEMORY 24 DL TOPS from GPU; High Perf. CPU CPU Complex: 2x Denver2 plus 4x Cortex-A57 Fully coherent HMP system; ARM V8 64-bit Pascal GPU – 5 th Gen GPU Architecture Custom acceleration for deep learning Separate memory for Tegra (CPU + iGPU) and dGPU Tegra (CPU + iGPU) to 8GB LPDDR4 (UMA): 50+GB/s Discrete GPU to 4GB GDDR5: 80+ GB/s 11
CORE DRIVE™ PX 2 CAPABILITIES RICH SW/HW CAPABILITIES Binary Compatibility Prepared to handle many use-cases Train on DevBox /DIGITS; Inference on DRIVE™ PX 2 Develop on PC (CUDA); Run on DRIVE™ PX 2 Develop on another NVIDIA embedded platform; Run on DRIVE™ PX 2 … By making the experience across homogeneous Unified tool chain (across DevBox , DRIVE™ PX2 and other platforms) Multimedia APIs aligned across all platforms (Linux) Full Open GL (not just ES) Ubuntu Unity desktop CUDA/CuDNN/DL Frameworks … 12
TOOLS & ECOSYSTEM RICH SW/HW CAPABILITIES Data Logging 2x 10 Gigabit Ethernet Uncompressed Camera stream 2x HDMI (4K/60) as Data pipe GMSL GMSL Camera Optional: for uncompressed video Bypass 12x Camera 12x Camera 10 Gigabit Ethernet Gigabit Ethernet DRIVE™ PX 2 Optional: for uncompressed video Data Logger ECU debug data Other Sensors & GPS (LiDAR/Radar) CAN/FlexRay Vehicle Information 13 Other Sensors
TOOLS & ECOSYSTEM RICH SW/HW CAPABILITIES CAN Bus configuration – EasyCAN* User can define 20 different CAN node IDs between Infineon Aurix and NVIDIA Tegra. No special tool-chain required. Messages defined through configuration files on Linux (running on Tegra A). E.g. one of the 20. CAN Channel F (on Aurix): Standard Can Id (0x244) - forwarded to Tegra B 14 * Provided by NVIDIA Partner Elektrobit
TOOLS & ECOSYSTEM RICH SW/HW CAPABILITIES Exquisite Developer Tools Graphics Debugger, System Profiler Multi Process, Multi Node, Multi GPU CUDA tools 15
TOOLS & ECOSYSTEM RICH SW/HW CAPABILITIES DRIVEWORKS Tools and building blocks to create your autonomous driving applications. Object detection, map localization, path planning and visualization. 16
TOOLS & ECOSYSTEM PRODUCT PARTNERSHIPS Camera, LiDAR and more USB/GigE Cameras LiDAR Image Sensors Hardware Reference board partners SDKs and frameworks from Partners 17
SPRINGBOARDS TO PRODUCTION FUNCTIONAL SAFETY SEooC & Work Products for DRIVE™ PX 2 Being developed as SEooC for autonomous driving Developing work products Functional and Technical Safety Concept – Scheme to achieve certain Safety Integrity Level for a particular use-case. System Design Specification – Functional details of HW / SW components & interfaces that are compliant to various standards (i.e. ISO26262) Hardware-Software Interface Specification – Register level details of HW/SW components Safety Manual – Safety architecture (e.g. RTOS/AUTOSAR) and list of all safety mechanisms and instructions to use them (e.g. DRAM ECC) FMEDA - Documentation of random failure metrics 18
SPRINGBOARDS TO PRODUCTION FUNCTIONAL SAFETY SW architecture Certified Hypervisor and foundation partitions Tegra/Pascal HW Safety OS (CPU Complex, SCE, Safety MCU) Safety supervision framework Tegra L1SS Safety diagnostics libraries IPC Safety AUTOSAR Stack (SCE, Safety MCU) L2SS WD STLib Applications, Middleware and Libraries SPI nSAFE L3SS STLib WD 19
SPRINGBOARDS TO PRODUCTION AUTOMOTIVE SW AUTOSAR DRIVE™ PX comes integrated with AUTOSAR 4.x-compliant software suite from Elektrobit Running on NVIDIA Tegra and Infineon AURIX™ 32 -bit TriCore ™ microcontroller Integrated Linux and AUTOSAR apps Functionality to monitor & redundancy management. IPC in safe/reliable execution environment. Also available reference stacks from other major AUTOSAR solution providers. 20
DRIVE™ PX 2 Closing remarks DRIVE™ PX 2: AI Supercomputer for Self Driving Cars Built for application development, rapid embedded prototyping and to help migrate to series production. Delivers powerful I/O and processing capabilities, rapidly expanding product ecosystem and several means to shorten the path to production. 21
REFERENCES DRIVE™ PX 2 Launch at CES https://youtu.be/C_8MZZ2TZUk (View Part 1 through 9) Deep Learning for Cars (GTC 2016) https://youtu.be/RVmV9SXJeBg & https://youtu.be/YuyT2SDcYrU NVIDIA Self-Driving Car https://youtu.be/5Wr_6Wk0DmY Roborace https://youtu.be/KN_qostst-s 22
April 4-7, 2016 | Silicon Valley THANK YOU JOIN THE NVIDIA DEVELOPER PROGRAM AT developer.nvidia.com/join
Recommend
More recommend
