Chinese ADAS and autonomous driving market to reach RMB42.6 billion in 2021
The advanced driver-assistance systems (ADAS) and autonomous driving market in China was worth about RMB5.9 billion in 2017 and is expected to reach RMB42.6 billion in 2021, growing at an AAGR of approximately 67%. Automotive vision, millimeter wave (MMW) radar, and ADAS are the market segments that are developing first, with the MMW radar market enjoying an impressive growth rate, closely followed by low-speed autonomous driving. The LiDAR (light detection and ranging), commercial-vehicle autonomous driving, and passenger-car autonomous driving markets are lagging behind.
The report, “ADAS and Autonomous Driving Industry Chain Report 2018 (I) - Computing Platform and System Architecture,” presented by ReportsnReports, presents a forecast for the market and discusses the strategy of ADAS and autonomous driving for carmakers including Geely, GM, SAIC, Dongfeng, Great Wall, GAC, Chang'an, NIO, Xpeng, and BYTON. It also addresses the software architecture of ADAS and autonomous driving, including AUTOSAR Classic and Adaptive, ROS 2.0, and QNX; as well as hardware architecture, including automotive Ethernet, TSN, Ethernet switch and gateway, and domain controller. Safety certification of ADAS and autonomous driving, including ISO26262 and AEC-Q100, is covered. Processor firms, including NXP, Renesas, Texas Instruments, Mobileye, Nvidia, Ambarella, Infineon, and ARM, are discussed as well.
As the automobile enters an era of ADAS and autonomous driving, product iteration increases and lifecycle of products is shortened. The automotive market is far smaller than the consumer electronics market but sees more complex designs and higher design and production costs than those in the consumer electronics market. For this reason, automotive ADAS and autonomous driving processor developers are faced with higher risks, which means that adequate financial and human resources are required to support the development. Globally, only a few enterprises, such as NXP and Renesas, are capable of developing a series of ADAS and autonomous driving processors.
As far as safety certification, autonomous driving chips must attain ASIL B at least, a level only Renesas R-CAR H3 has reached to this point. As GPU is a universal design and not a car-dedicated design, it is difficult to reach the certified safety level of ISO26262 from the point of design. The certification cycle of ASIL is up to two to four years.
Reliability, precision, and functionality of stereo cameras are well above those of mono cameras, but as the stereo camera must use field-programmable gate array (FPGA) technology, it is costlier. High costs limit the application of the stereo camera to luxury cars. However, with the emergence of Renesas and NXP hardcore stereo processors, the stereo camera will be widely used in the ADAS and autonomous driving field, expanding from luxury to mid-range models.
With an explosive growth in data transmission, automotive Ethernet will become a standard configuration of the automobile. The Ethernet gateway or Ethernet switch is indispensable to autonomous driving.
Autosar will act as a standard configuration in ADAS and the autonomous driving field.
In the area of CNN/DNN graphics machine learning, GPU is most suitable when data is irrelevant to sequence. Nvidia GPU can be used in multiple fields other than automotive, and shipments are far higher than those of the automotive application-specific integrated circuit (ASIC), as the former present superior cost performance. TPU is faster and uses less power (only 10% of that of GPU) at the expense of the precision of computation.
Concerning RNN/LSTM/reinforcement learning sequence-related machine learning, FPGA has distinct advantages, particularly in power consumption, consuming less than one-fifth of GPU under the same performance. However, high-performance FPGA is costly. FPGA can also process graphics machine learning and improve performance.
ASIC stands out by performance-to-power consumption ratio but has shortcomings of long development cycle, the highest development cost, and the poorest flexibility. The unit price will be high or firms will incur losses if the shipments are small (at least annual shipments of 120 million units if 7-nanometer process is employed). Most ASICs for deep-learning graphics machine learning are similar to TPU.
Power consumption and cost performance are crucial in the in-vehicle field. GPU is a winner in graphic machine learning. However, as algorithms are constantly improved, the low requirements on the precision of computation, and low power consumption will ensure a place for FPGA in graphics machine learning. FPGA has significant advantages in sequence machine learning.
Autonomous driving can be divided into two types, one represented by Waymo, which has solved most of the problems concerning environmental perception and concentrates on behavior decision-making with computing architecture of CPU+FPGA (usually Intel Xeon 12-core and above CPU plus Altera or Xilinx's FPGA); the other represented by Mobileye, which has not solved all problems involving environmental perception and concentrates on it with computing architecture of CPU+GPU/ASIC.
CPU+GPU will be the mainstream in the short run, but CPU+FPGA/ASIC may dominate in the long term, largely due to continuous decline in the precision of computation of graphics because of improvement in algorithms and performance of sensors (LiDAR in particular), which is conducive to FPGA, while it is difficult for the power consumption of GPU to fall. It is easier for FPGA to meet car-grade requirements.
In the chip contract manufacturing field, TSMC has won all 7-nanometer chip orders, including A12 exclusively provided for Apple. This is the first time that TSMC overtook Intel to become the vendor with the most advanced semiconductor manufacturing process, a must in the production of digital logic chips whose computing capability is necessary for artificial intelligence and autonomous driving.