According to CCTV news, Tesla CEO Elon Musk arrived in Beijing on the afternoon of the 28th at the invitation of the China Council for the Promotion of Trade. What is Musk's intention in this visit to China?

Global Network quoted the latest Reuters report as saying that this was “an unexpected visit” by Musk to China.

Notably, a week ago, a user asked Musk on a social platform, “I'm looking forward to Tesla FSD entering China as soon as possible. When will this happen? When will the HW3.0 system be able to use 3D modeling images? When will the new reversing assist function be rolled out to Tesla HW3.0 owners in China?”

At the time, Musk responded, “This may happen soon.”

What is the current level of FSD?

Just this month, Tesla “took action” on the FSD price twice, lowering the price to attract users.

A week ago, Tesla announced that it had cut its FSD system package price by about one-third from 12,000 US dollars to 8,000 US dollars after lowering the price of models in the Chinese and US markets. Earlier, on April 13, Tesla announced that the FSD (supervised) monthly subscription price had dropped from 199 US dollars to 99 US dollars, directly “falling” the subscription price.

Musk expects Tesla's FSD test mileage to reach 6 billion miles to meet the requirements of global regulators. According to data released by Tesla on April 6, the actual test mileage of the FSD was 1 billion miles.

Although it is still far from the target, Musk announced at the beginning of this month that it will launch Robotaxi and commercialize it on August 8 this year — this also means that Tesla will need more FSD users in a short period of time and increase road test mileage.

Currently, Tesla FSD has been upgraded to the V12.3 end-to-end version. It no longer relies on traditional rule code, but completely relies on neural networks and data drivers.

According to CITIC Securities's March 28 report, the core benefit of “end-to-end” is that the key to model iteration has changed from “engineers” to “data and computing power” that is easier to scale, so training efficiency and performance limits are expected to be significantly improved. The performance potential shown in this solution is expected to significantly improve the level of takeover of autonomous driving systems, thereby achieving truly undisputed L3 capabilities (for example, reaching 1 takeover per week).

Judging from user feedback, V12's performance is indeed outstanding: According to feedback from North American test users participating in FSD Beta V12, compared to V11, the number of interventions with the V12 version was significantly reduced, and “more silkier” and “more personable” are also common reviews.

What does FSD need to enter China?

Last year, it was reported in the media that Tesla is building its first data center. Its data center business can localize data storage, and all data generated by selling vehicles in the mainland China market will be stored within China. Furthermore, Tesla is setting up a local operation team of about 20 people in China to promote FSD's launch in the Chinese market; it is also preparing to establish a data labeling team of hundreds of people in China to provide manual labeling samples for training FSD algorithms.

If FSD is implemented in China, data collection, model training, and computing power centers are seen as the key.

CICC once said that if FSD is promoted in China, the complete data loop needs to be in the country, and preparations need to be made in terms of data collection and model training to consider whether the BEV sensing solution can be well adapted for localization. In China, FSD map surveying may require cooperation with Chinese manufacturers, and Tesla needs to build a local computing center and set up a local algorithm adaptation team.

CITIC Securities also pointed out, first, that the map data is highly sensitive and closely related to national security. It is still unknown how Tesla will obtain relevant qualifications, or enter into joint ventures with Chinese companies or conduct pilot projects in Shanghai, but it is not easy to expand on a large scale in a short period of time.

Second, the pure vision route relies more on extensive data collection and model training for image processing, so Tesla built its own supercomputing center Dojo in the US and developed its own D1 chip to greatly improve FSD training efficiency. The brokerage firm believes that in China, even if Tesla obtains data collection qualifications, it is likely that it will only be able to train models locally in China. This means that to achieve the same training efficiency as the US, Tesla needs to establish a supercomputing center similar to Dojo in China, which requires a certain development cycle and cost.

What is the impact of FSD entering China?

What was the impact of Tesla FSD after entering China? Combining the research and judgment of various brokerage firms, it may greatly enhance consumer awareness, promote the penetration of advanced intelligent driving in China, and trigger the “pangasius effect” in the field of intelligent driving in China.

For example, CITIC Securities said that if Tesla's “ultimate cost-effective hardware+high-margin FSD software” model is implemented on a large scale in China, it may be similar to the “ChatGPT moment” for smart cars; whether in terms of incremental demand or business model, it means new possibilities for current domestic car companies and supply chains.

Guotai Junan's March 26 report also pointed out that if FSD successfully enters China, it will become a strong competitor for domestic car companies' autonomous driving solutions, and the competition for domestic autonomous driving highland will become more intense, which is conducive to the flourishing development of the autonomous driving industry.

edit/ruby