Tesla earnings conference call record: the compact model is not considered for the time being, and automatic driving is the focus


Tencent technology, January 26 – Tesla (NASDAQ: tsla) released its fourth quarter financial statements as of December 31, 2021 on Thursday. According to the financial report, Tesla’s total revenue in the fourth quarter was US $17.719 billion, a year-on-year increase of 65%; The net profit attributable to Tesla was $2.321 billion, a year-on-year increase of 760%.
After the financial report was released, chief executive Elon Musk, chief financial officer Zach kirkhorn, senior vice president of Engineering drew Baglino and investor relations director Martin viecha attended the financial report teleconference and answered analysts’ questions on the spot.
The following is the actual record of the earnings conference call:
Analyst: what’s the latest news about the $25000 compact car?
Elon Musk: we don’t have a $25000 compact car right now. We’re very busy right now. You shouldn’t ask this question. What’s more important is when the vehicle can become autonomous. In this way, the transportation cost will be reduced four or five times.
Analyst: since we talked about the product development path today, what is your view on household refrigeration and heating from the perspective of accelerating the transformation of sustainable energy? In addition, what role will Tesla’s HVAC {and heat pump technologies play?
Drew bagrino: from the perspective of mission, our heat pump technology is consistent with the transformation of sustainable energy. Imagine that using electric heat pump to replace natural gas and water in indoor heating system will offset 80% of the total energy saving of solar roof, which has a considerable influence. We learned a lot about how to develop a heat pump that can work reliably under all environmental conditions, and we are very excited about it.
Elon Musk: we integrate electric vehicles, charging piles, solar energy, energy storage, hot water and HVAC into a very compact package product, which will be a very attractive solution for large customers. It looks great, but there is no such product yet.
Lars moravy, vice president of vehicle engineering: the integration of these systems in the house and in the car is actually the same thing, but it is much more difficult in the car. It is limited by mass, volume and energy, but the main house is very simple. As we integrate more modules, it will naturally extend to HVAC and hot water systems. We will certainly do this and integrate better than other competitors.
Elon Musk: when you go home, everything in the house can start heating or cooling, like this. We think it will overturn the tradition, but we have too many important things to do. We will do it, but we won’t have a specific time plan.
Lars moravi: everyone should do it. It’s very good.
Analyst: will you consider dividing the authorization of fully automatic driving (FSD) package into permanent authorization and periodic authorization (e.g. vehicle life is the term), or will you authorize a higher level of permanent or periodic authorization for commercial vehicles? Is it possible for permanent authorization to be directed at the owner or business owner rather than the vehicle itself?
Elon Musk: No, it sounds too complicated. We’ll just think about how to make the cost per mile lower.
Analyst: will Dojo (Tesla neural network supercomputer project) be available in the summer of 2022? Do you have any challenges? If you want to drive in a city like New York, is Dojo necessary to make fully automatic driving (FSD) run better? In addition, where will Tesla humanoid robot be used for the first time? In your factory?
Elon Musk: Yes, Dojo will come in handy this summer, focusing on when it is more competitive in training data than GPU clusters. I want the fully autonomous driving (FSD) team to tell me they want to switch to dojo. I can’t say it will be 100% successful, but it should be achieved next year, when Dojo will do well. Dojo is not necessary for fully automatic driving (FSD), but it can optimize the cost of training massive videos. The reduction of cost also means the enhancement of performance. If we can train the model faster, the project will progress faster. Not all data can be put on the GPU, and some need to be zeroed. If dojo is competitive enough, we will provide it to other enterprises that want to conduct neural network training. This is actually an optimized neural network training system, especially for video data. We encourage those who think dojo is an interesting project to join us.
Tesla humanoid robot is called Optimus. If we can’t find a place to use it, others must have no reason to use it. We’ll let it transport parts in the factory.
Continuously updating