China Electronics：4D mm-wave radar: the standard sensor solution for better cost-to-performance that could advance self-driving

The gist

·?Industry catalyst and rising profile: Tesla might bring back imaging radar

·?Ideal for mass production: wide efficiencies; huge array; special processors

·?Total value chain growth; global 4D radar market of USD16bn in 2030E

?

Industry catalyst and rising profile: Tesla might bring back imaging radar

Tesla’s self-driving sensor solutions have gone through three main phases of development: (1) the fusion of multiple sensors; (2) pure vision; and (3) 4D mm-wave radar plus vision. It could finally adopt the imaging radar for its advantages of pitch angle ability, high resolution and lower cost than lidar, which mitigate the shortcomings of a pure vision solution. An earlier radar used by Tesla could not process height. Weighing overall performance and cost considerations, we believe a better solution for HW4 is the fusion of cameras and 4D mm-wave radar. Aside from Tesla, proponents of 4D mm-wave radar include Chinese and foreign car companies like BMW, GM and Changan; some have made deliveries while others plan to do so in 2023.

?

4D mm-wave radar is the optimal solution for mass production: wide-spectrum performance improvements; ultra-large arrays and dedicated processors

4D refers to the comprehensive analysis of distance, orientation, speed and height of a detected object with ultra-high resolution. The three main tech routes to 4D imaging are: (1) cascading; (2) single-chip integration and (3) algorithmic virtual aperture. Cascading technology is the mostly widely used currently. Arbe’s 4D mm-wave radar, which integrates a large array of RF chipsets and dedicated processing chips to offer 2,304 virtual channels, has a mass production price range of USD100-150, currently the most cost-effective solution. We think Tesla might adopt the Arbe solution. 4D mm-wave radar would benefit from the trend of process integration in the future as its cost continues to decline. We estimate a cost decline trend of 20/12/2% in 2023/25/30E as it is deployed in various types of vehicles. The downstream application scenarios are diverse. We project the application market for autonomous driving and non-autonomous driving could reach USD11bn and USD17bn in 2025E, respectively.

?

Application advantages: standard self-driving solution for software-defined cars

We believe 4D mm-wave radar is the most economical and stable standard solution for in-vehicle radar. We believe it could replace the traditional mm-wave radar and low-performance lidar, while complementing high-performance lidar. It could see more frequent usage in certain models. Compared to traditional solutions, 4D mm-wave radar solves limitations in static identification, lateral movement detection, height identification, adjacent object discrimination and hidden vehicle detection. Compared to lidar, 4D mm-wave radar has several performance indicators that come close, while overcoming making lidar’s shortfalls in speed and distance detection, as well as adverse weather and environmental factors. Yet, its cost is only one-tenth of lidar’s. When it comes to software-defined vehicles, we believe 4D mm-wave radar provides richer data and more reliable solutions in the support of autonomous driving.

?

Total spectrum value chain growth; global 4D radar market of USD16bn in 2030E

The mm-wave radar value chain includes:

·??Upstream: RF MMIC chips, high-frequency PCBs, processing chips, back-end algorithms

·??Midstream: finished 4D mm-wave radar manufacturers

·??Downstream: OEMs.

?

In the upstream market, software algorithm barriers account for 50% of the total market, followed by radio frequency chips, processor chips and high-frequency PCBs. With the increasing penetration of 4D products, the value of the entire industry chain could rise. Our estimates project a global automotive 4D mm-wave radar market of USD318m/ 1.59bn/16.16bn in 2023/25/30E.

?

4D imaging technology could transform competitive dynamics in entire industry

Among whole-system suppliers, traditional tier-1 manufacturers generally adopt cascade technology and lead in the mass production of 4D products, while newer entrants might rely on dedicated chipsets and virtual aperture solutions to enable lane-changing and overtaking. Products by Arbe and Mobileye are highly competitive for offering multiple channels. Many Chinese manufacturers have released 4D mm-wave radar prototypes in cooperation with Chinese automakers. Major whole-system producers include Jingwei Hirain, Weifu High-tech, Huawei, Chuhang Technology and United Optech. Upstream chips mainly comprise RF chips and processors, and also dedicated chips that integrate the two solutions. Major companies in this segment are Texas Instruments, Xilinx and NXP. As 4D computing power demand increases, the solutions have evolved into two major ones: FPGA and DSP. China’s Calterah leads in the development and design of mm-wave radar chips with CMOS technology. In upstream PCBs, major players include Rogers, Isola, Panasonic, as well as domestic names like WUS Printed Circuit and Shengyi Technology (Shengyi Electronics).

?

Stock ideas (names in bullet points)

We suggest that investors monitor these 4D mm-wave radar players:

·?Chip packaging: JCET Group (600584 CH, BUY)

·?PCBs: WUS Printed Circuit (Kunshan) (002463 CH, ACCUMULATE) (joint coverage with the communications desk), Shengyi Electronics (688183 CH, BUY)

·??Whole-system producers: Weifu High-Technology Group (000581 CH, BUY), Beijing Jingwei Hirain Technologies (688326 CH, BUY) (joint coverage with the autos desk), Huayu Automotive Systems (600741 CH, BUY), Shanghai Baolong Automotive (603197 CH, BUY) (joint coverage with the autos desk).

?

Not-rated market players

Olympic Circuit Technology (603920 CH), Huizhou Speed Wireless Technology (300322 CH), Union Optech (300691 CH).

?

Risks include: lower mass production of new technologies than expected; intensifying competition; and lower-than-expected NEV penetration.

Disclosure

?This publication is distributed in Hong Kong by TFI Asset Management Limited (“TFIAM”), which is regulated by the Securities and Futures Commission of Hong Kong (“SFC”) and is licensed for the conduct of Regulated Activity Type 4 (Advising on Securities), Type 5 (Advising on Futures Contracts), and Type 9 (Asset Management) of Part V of the Securities and Futures Ordinance (Cap.571) (“SFO”) with Central Entity ASF056.

Information contained herein is based on sources that TFIAM believed to be accurate. TFIAM does not bear responsibility for any loss occasioned by reliance placed upon the contents hereof. The information is for informative purposes only and is not intended to or create/induce the creation of any binding legal relations. The information provided do not constitute investment advice,solicitation, purchase or sell any investment product(s). Investments are subject to investment risks including possible loss of the principal amount invested. You should consult your Financial Advisor for investment advice based on your investment experience, financial situation, any of your particular needs and risk preference.

TFIAM and/or its representatives may have positions/ interests in relevant investment products. TFIAM and/or its affiliates may from time to time provide services for, or solicit services or other business from, any company mentioned in this report. The above information is owned by TFIAM and protected by copyright and intellectual property Laws. It may not be reproduced,distributed or published for any purpose without prior written consent from TFIAM.