NVIDIA Corporation Business Report FY ended Jan. 2018
Financial Overview |
(in USD millions) |
| FY ended Jan. 31, 2018 |
FY ended Jan. 31, 2017 |
Rate of change (%) |
Factors | |
| Sales revenue | 9,714 | 6,910 | 41 | Each segment contributed to strong growth. GPUs for gaming is the primary contributor to sales revenue. |
| Net income | 3,047 | 1,666 | 83 | - |
| Sales by division | ||||
| Automotive | 558 | 487 | 15 | Agreements for the development of infotainment modules, DRIVE PX, and autonomous driving were the primary contributors. |
NVIDIA DRIVE AGX autonomous driving platform
NVIDIA has been providing GPU (Graphic Processing Unit), such as graphics boards, for gaming and PC applications, and has announced its Xavier SoC (System on Chip) that uses a combination of CPU, GPU, and Tensor Cores (a collection of multidimensional array calculators optimal for deep learning) that is based on its current Volta GPU architecture.
The NVIDIA DRIVE AGX is a scalable, open autonomous vehicle computing platform that uses the Xavier SOC.
The NVIDIA DRIVE AGX platform currently has two types of products in its lineup, of which the most popular product is the top-end NVIDIA DRIVE AGX Pegasus AI computer with an architecture built on two Xavier SoCs and two next-generation Tensor Core GPUs.
NVIDIA has announced that its NVIDIA DRIVE AGX Pegasus system uses Tensor Cores (a collection of multidimensional array calculators optimal for deep learning) with a computing capacity of 320 TOPS (Tera Operations per Second). The NVIDIA DRIVE AGX Pegasus platform demonstrates its computing power by processing data from various on-board sensors such as cameras, radar, and LiDAR equipped on an actual autonomous vehicle as well as the processing of position information and route prediction information.
NVIDIA provides the NVIDIA DRIVE AGX autonomous vehicle computing platform in combination with this NVIDIA DRIVE AGX Pegasus system and software.
A key component to the function diversity of the NVIDIA DRIVE AGX autonomous vehicle computing platform is software. In addition to self-driving capabilities, DRIVE Software enables in-vehicle functionalities, such as DRIVE IX for driver monitoring and user experience and DRIVE AR for visualization of the car's perception of its environment. The NVIDIA DRIVE Software suite integrates both the DRIVE AV and DRIVE IX software stacks are used in the NVIDIA DRIVE AGX for self-driving vehicles. At its core is DRIVE OS, which delivers a safe, secure, real-time hypervisor and other core capabilities and services for autonomous driving applications.
For autonomous vehicles, the company offers its supercomputer for data centers, and the NVIDIA DGX workstation deep learning system. As never before, the NVIDIA DGX station provides the advanced processing capability of the data required for AI-based deep learning.
Currently more than 450 automakers, suppliers, sensor companies, mapping companies and others companies around the world are using the NVIDIA DRIVE AGX platform, of which 55 companies are headquartered in Japan.
DRIVE Constellation simulation platform for automated driving systems
In addition, NVIDIA also provides a running simulation environment to realize the test driving distances required to develop automated driving vehicles.
The simulation service called DRIVE Constellation processes simulated virtual data in the same way as actual driving tests using the aforementioned NVIDIA DRIVE AGX Pegasus system.
The simulation platform virtually creates data from various sensors such as cameras, radars, LiDAR equipped on actual vehicles using riders and the like obtained by actual traveling vehicles, using DRIVE Sim software. As a result, it is possible to use simulation data to complement the road testing required to develop automated vehicles, and is an efficient means for the development of actual driving scenarios.
The company also offers its DRIVE AR software for simulation, which uses augmented reality (AR) to simulate the car’s perception of its driving environment.
In September of 2018, NVIDIA announced that DRIVE Constellation and DRIVE Sim will be released to its simulation partner companies. The objective of opening up the DRIVE Constellation platform is to learn from each other through collaboration to improve the depth and flexibility of the platform, and to improve the effectiveness of testing and verification.
>>> Click here for reports on NVIDIA DRIVE AGX
Business Partnerships
-As of 2018, the company has partnered with more than 320 OEMs, mapping companies, and university research institutions, etc.
| Collaborating company | Content | ||
|
OEM |
|||
| Isuzu Motors | Development of its next generation trucks equipped with advanced driver-assistance features and autonomous driving technologies using the NVIDIA DRIVE AGX platform | ||
| Yamaha Motor Co. | Adopting the Jetson AGX Xavier development system to power its upcoming lineup of autonomous machines, with plans introduce unmanned agriculture vehicles | ||
| Daimler | Adopted NVIDIA’s AI platform as part of its control unit network for automated driving and unmanned operation in urban areas | ||
| VW | Cooperating in the field of deep learning, and adopting NVIDIA’s AI technology for automated driving of its new I. D. BUZZ model and a mini bus | ||
| Toyota | Plans to use the NVIDIA DRIVE AGX Xavier processor as the AI brain of its vehicles installed with the NVIDIA DRIVE PX autonomous driving system in its vehicles to be released from 2020. | ||
| SUBARU | Targeting the introduction of vehicles with Level 2 or higher autonomous capability for highway driving by 2024, Subaru will incorporate NVIDIA’s AI technology to realize surrounding driving environment recognition, path planning, multiplexing, and intelligent vehicles. | ||
| Audi | Developed the Audi AI Traffic Jam, which is the world’s first system that enables Level 3 autonomous driving | ||
|
Suppliers |
|||
| Bosch | Adopted NVIDIA’s AI platform as part of its control unit network for autonomous driving and unmanned vehicles in urban areas using the NVIDIA DRIVE PX computer installed in its automated and autonomous driving systems | ||
| Pioneer | Made its 3D LiDAR sensors compatible with the NVIDIA DriveWorks software development kit for autonomous driving | ||
| Continental | Collaborating to develop self-driving vehicle solutions based on the NVIDIA AI car computer and the NVIDIA Drive platform for autonomous vehicle development | ||
| Infineon Technologies | The company's AURIX TC3xx microcontroller is expected to be adopted for NVIDIA’s DRIVE Pegasus on-board computing platform for Level 5 automated vehicles | ||
| ZF | Collaborating on the development of AI to be equipped in autonomous vehicles for mass production. NVIDIA provides its DRIVE PX solution to support image processing technology and driver-assistance systems jointly developed by ZF and HELLA. Cooperating on the development of AI systems as well as autonomous driving and automated driving systems for segments such as the transport industry as well as passenger vehicles and trucks. |
||
| HELLA | Collaboration in the development of AI to be equipped in passenger cars for mass production. NVIDIA provides its DRIVE PX solution to support image processing technology and driver-assistance systems jointly developed by ZF and HELLA. | ||
|
Start-Ups |
|||
| Uber | Adopted NVIDIA’s AI technology system for its fleet of autonomous vehicles | ||
| Uniti | Beginning with the NVIDIA DRIVE PX, the company's will equip its Uniti One pure EV with all of the equipment necessary for autonomous vehicles | ||
| Baidu | Collaboration in a wide range of fields that utilize AI for autonomous vehicles and cloud data centers. Baidu has adopted the NVIDIA DRIVE PX Platform and is cooperating with a number of major Chinese automobile manufacturers to develop autonomous vehicles. | ||
| Zenity (Volvo & Autoliv joint venture) |
The JV will use NVIDIA's AI on-board computing platform to develop Level 4 autonomous vehicles by 2021, | ||
|
Maps |
|||
| Zenrin | Collaborative research on high definition map solutions for autonomous vehicles. Performs data processing using NVIDIA DRIVE PX 2 AI and NVIDIA DRIVE Works in mobile mapping vehicles, and the NVIDIA GPU and NVIDIA MapWorks for its data centers. | ||
| TomTom | Uses a combination of the NVIDIA DRIVE PX2 and AI for "cloud to car" services which can update cloud-based HD maps in real time. | ||
R&D Expenditure
|
(in USD millions) |
| FY ended Jan. 31, 2018 | FY ended Jan. 31, 2017 | FY ended Jan. 31, 2016 | |
| Total | 1,797 | 1,463 | 1,331 |
Product development
On September 26, 2017, the company announced that its DRIVE hardware and software solutions have been adopted by 145 autonomous driving startup companies. On the same day at the NVIDIA GTC conference in Beijing, the company also announced its Tensor RT 3.0 platform. Combined with the company's GPUs, the Tensor RT is a platform for high-performance AI-based deep learning that can be deployed quickly and efficiently for AI inference service to develop self-driving cars and robots.
In October 2017, the latest version of the DRIVE PX AI computer for its automated driving platform was announced. The code name for the platform is Pegasus and unit is about the size of a license plate. It was designed for Level 5 autonomated vehicles such as robo-taxis. More than 25 out of the 225 affiliated companies are developing fully automated robo-taxis equipped with NVIDIA’s CUDA GPU and the DRIVE PX Pegasus AI car computer, with plans to introduce the system in vehicles of its partner automakers in the second half of 2018.
On-board information
In September 2017, Audi announced the first production car in the world to have been developed for Level 3 highly automated driving, the Audi A8, which it exhibited at the 2017 Frankfurt Motor Show (IAA), utilizing NVIDIA technology and equipped with various piloted driving functions. One of these functions, the AI Traffic Jam Pilot uses the zFAS driver-assistance system equipped with NVIDIA products to enable autonomous driving at speeds of up to 60 km/h.
Also in September 2017, Porsche announced that it is providing navigation and infotainment services on its Cayenne Turbo model equipped with a 2.3-inch full HD touch panel using NVIDIA’s Tegra processor.
In December 2017, NVIDIA announced information on some of its technologies that is being adopted for some of the models exhibited at the Los Angeles Motor Show.
- Tesla announced that it is using the NVIDIA DRIVE PX AI computing platform to run its Telsa Autopilot autonomous driving system that it developed using large neural networks to manage the vision, sonar, and radar sensors installed in all Tesla models.
- Porsche's new Panamera Turbo S E-Hybrid Sport Turismo is equipped with an infotainment system using NVIDIA’s systems to provide functions such as vivid graphics, navigation, and real-time traffic information.