ChinAI #265: The Race to Become China's No. 1 Autonomous Driving Chipmaker

Horizon Robotics vs. Black Sesame Intelligence

Greetings from a world where…

tricornes are the new beanies

…As always, the searchable archive of all past issues is here. Please please subscribe here to support ChinAI under a Guardian/Wikipedia-style tipping model (everyone gets the same content but those who can pay support access for all AND compensation for awesome ChinAI contributors).

Feature Translation: Horizon Robotics vs Black Sesame

Context: On March 22, 2024, Black Sesame Intelligence [黑芝麻智能] renewed its application to list on the Hong Kong stock exchange. A mere four days later, Horizon Robotics [地平线机器人] also submitted its own prospectus to list on the same exchange. Both are unicorns (with valuations well above $1 billion), both are registered in the Cayman Islands (read: have lots of foreign investment), both have their HQs in China (Horizin in Beijing; Black Sesame in Wuhan), and both are competing for the same market: making system-on-chips (SoC) that power autonomous cars and advanced driving assistance systems. This week’s feature translation, an analysis of the IPO documents provided by both companies (link to original Caijing E-Law piece), uses the head-to-head competition between Horizon and Black Sesame to shed light on China’s autonomous driving chip industry.

Key Takeaways: Horizon and Black Sesame are the two leading Chinese manufacturers of high-compute SoC for autonomous driving, though Nvidia still dominates the market.

• Per the data cited by Black Sesame’s prospectus, Nvidia captured 81.6% of China’s market for this segment in 2022. Horizon came in second with a market share of 6.7%, while Black Sesame ranked third with a 5.2% market share (see table below)

• As is the case with many types of chips, leading firms have so many cumulative advantages. Caijing E-Law cited one industry analyst who stated that if a new company emerged to compete with Horizon, the gap may be three years or even longer, due to upfront R&D costs, long development + deployment cycles, and constant technical iteration.

How will Horizon, Black Sesame, and other emerging Chinese companies in this domain try to overtake Nvidia?

• They are betting that China’s adoption of advanced driving assistant systems and autonomous driving will continue to accelerate. One stat from the article: By 2030, more than 80% of autonomous driving technology in China’s passenger cars will be high-end solutions (higher levels of autonomy that require chips like the ones Horizon and Black Sesame specialize in).

• One important line from Black Sesame’s prospectus: “Chinese automotive OEMs are actively cooperating with local autonomous driving SoC suppliers to establish a secure and reliable local supply chain. Domestic autonomous driving SoC suppliers are likely to be more favored by Chinese automotive OEMs in the Chinese market as geographical proximity allows domestic companies to respond to the latest market demands and customer demands in a more efficient manner (emphasis mine).” The localization rate for China’s chips in general is projected to increase to 19% by 2025.

• According to one estimate by the China Association of Automobile Manufacturers, more advanced smart cars will require 3,000 chips per vehicle.

Dig deeper:

• Software matters even when the vertical is hardware. Horizon and Black Sesame have two main revenue streams: the SoC-based solutions and algorithm-based solutions, which refers to the licensing of software and development tool chains to customers. Horizon’s revenue from licensing increased from 43% in 2021 to 62% in 2023; by comparison, Black Sesame’s proportion of revenue from software licensing declined from 54 to 26.5 percent over this same period.

• See this previous ChinAI issue on Chinese automotive chips, with some interesting details about China’s standards system to verify chip performance in actual vehicles.

FULL TRANSLATION: Horizon Robotics VS Black Sesame, who will win the title of “China’s No. 1 Intelligent Driving Chip Stock”?

ChinAI Links (Four to Forward)

Should-read: NEC Silicon Valley - the Lab that Dragged Chinese Companies into the AI Era (ChinAI #220)

Both Yu Kai, Horizon’s CEO, and Wei Xu, Horizon’s chief scientist, trace their roots to a fascinating wellspring: NEC Silicon Valley (the California outpost of the Japanese multinational electronics corporation). This April 2023 ChinAI issue reviews the history of this fascinating lab, which played a pivotal role in the careers of so many key figures in China’s AI ecosystem.

Should-read: 5 colleges lose Chinese flagship programs because of Defense Dept. cuts

Last week, Karin Fischer’s essential rundown of top stories in international education featured the DOD’s surprising closures of its Chinese flagship programs. Let’s do an exercise: If we took a fraction of all the proposed new DOD investment into quantum and put it into Chinese-language learning, how much safer would we be as a country? Five times? Ten times? H/t Joelle Brown for sharing.

Should-attend: State of AI Safety in China (Concordia AI Spring 2024 Report Launch Event)

This Wednesday (5/15) at 9AM, on a one-hour Zoom webinar, I’ll join the Concordia team to discuss their new report on the state of AI safety in China. Looking forward to hearing insights from fellow panelists: Matt Sheehan, Robert Trager, and Angela Zhang.

Should-read: The Teens Making Friends with AI Chatbots

From the Longreads summary, by Krista Stevens: “Growing up, you might have confessed your angst to a paper diary, a blog, or maybe Myspace. Now, as Jessica Lucas reports for The Verge, teens can turn to AI chatbots to pour out their woes. But what are the emotional and societal consequences of relying on bots for support? Are the kids going to be alright?”

Thank you for reading and engaging.

These are Jeff Ding's (sometimes) weekly translations of Chinese-language musings on AI and related topics. Jeff is an Assistant Professor of Political Science at George Washington University.

Check out the archive of all past issues here & please subscribe here to support ChinAI under a Guardian/Wikipedia-style tipping model (everyone gets the same content but those who can pay for a subscription will support access for all).

Also! Listen to narrations of the ChinAI Newsletter in podcast format here.

Any suggestions or feedback? Let me know at chinainewsletter@gmail.com or on Twitter at @jjding99