Buying autonomous electric vehicles will be just like buying a laptop – Embedded

Place your AD here


In the future, automakers like Mercedes and BMW will be selling vehicles with marketing like Nvidia inside or Mobileye inside, just like a laptop might be bought that says Intel inside today.
With the high dependence of cars on semiconductors, consumers could start caring more about what chips are inside before making decisions on which autonomous vehicle (AV) or electric vehicle (EV) to buy.  According to a new report, purchasing a new vehicle will soon feel like shopping for a new laptop.
Referring to the report from research firm IDTechEx, “Semiconductors for Autonomous and Electric Vehicles 2023-2033”, the company’s senior technology analyst James Jeff said, “In the future, automakers like Mercedes and BMW will be selling vehicles with marketing like Nvidia inside or Mobileye inside, just like a laptop might be bought that says Intel inside today. The automakers will become system integrators, like how Lenovo buys CPUs from Intel or AMD, memory from various suppliers, and screens from LG and put them together into its own packaging. Automakers will buy an electric powertrain from one company and an autonomous system from another and bundle it into their branded packaging.”
Indeed, when I interviewed Clipper Automotive in London earlier this year, I was told similarly by its co-founder and CEO Janosch Oppermann that the automotive industry essentially sources various different parts of a vehicle from different parts manufacturers and systems integrators and carries out a system level integration to develop their own vehicles.
Hence IDTechEx says that consumers will choose cars based on how long the battery will last, how nice the screens are, and what computer chip it comes with. The report analyzes trends for semiconductor content across eighteen component categories addressing advanced driver assistance systems (ADAS), automated driving, electrification, communications & infotainment. Automotive microcontroller (MCU) architectures are the leading component for generating semiconductor value within the vehicle.
The report indicates the value of MCUs within vehicles is going contribute to a wafer revenue CAGR of 9.4%, but much of the growth is going to be driven by growing semiconductor demand within ADAS, AVs and vehicle electrification. It adds, “Not only will these new components require additional MCUs, but the advanced and intensive processing undertaken in automated driving is seeing the adoption of more cutting-edge semiconductor technologies into the vehicle.”
The similarities to the laptop clearly would center around things like how long the battery lasts and what chips define the user experience.
When buying a laptop, for most people, it will be things like how long the battery will last, how nice the screens are, and what computer chip it comes with. With AVs and EVs, buyers’ selection criteria will no longer be based on the number of cylinders, horsepower, and miles per gallon. Instead how long the battery lasts will be an obvious deciding factor for EV buyers. And then it will be about things like in-cabin infotainment.
On the latter point, ‘screenification’ of car cabins over the past few years has become more common. Even the cheapest cars on the market are available with some kind of central display, while the top end of the market is heading towards pillar-to-pillar style screens in the front. With talk about software defined vehicles and the ability to deliver services with over-the-air updates, the screens simply help automakers with enabling more premium features in the vehicle and likely more subscriptions.
Like laptops, vehicles are also becoming gaming devices. Tesla for instance generated waves when it showed the Model S could run steam on its infotainment system, showing how Witcher 3 can be played when the vehicle is stationary. Mercedes went one step further; now that it has level 3 certification in both the US and Germany, it is possible for the driver to play video games while the car drives itself, up to a speed of 40mph. However, where Tesla is playing games one might expect of a modern gaming laptop, the Mercedes allows the play of Tetris, Sudoku, and some mobile gaming titles.
Anecdotally, we also saw this as a big trend at the FT Future of the Car Conference in London last month. Elena Boisseuil, the customer journey director at Nissan Motor Corporation, talked about how the screen will be a huge opportunity for the company. Niraj Seth, a vice president at Cognizant, said, referring to the car of the future, “It’s not just a form of transport.” In an open Q&A session, one developer on the Honda e highlighted how successful one of its screen apps was, contrary to their expectations that customers wouldn’t be interested.
While gaming in vehicles might be considered a gimmick or novelty, the IDTechEx report said that the real computation power in modern and future vehicles will be used to drive autonomous systems. Most computer chips around the vehicle, the kind that are used to open windows or adjust wing mirrors, use quite mature technologies. These chips have node sizes normally above 40nm (nanometers). Modern smartphones, on the other hand, have chips that use a 4nm process, with smaller processes offering more computing power. However, the coming chips that will power autonomous driving systems will be much closer to the cutting edge, using technologies in the 1-5nm region.
This is where laptops come back in. A high-performance gaming laptop might be using a graphics card from Nvidia, giants in the gaming industry. The graphics cards take information from the CPU and turn it into an image for the screen. One of the main computational tasks for an autonomous vehicle is to take data from each of the cameras, radar, lidar, and so on, and turn it into a 3D map of the environment and identify all the vehicles and all the people around the vehicle.
These two processes are rather similar, and it means that Nvidia has been able to expand into the automotive space, offering top-end, high-performance computing for autonomous applications. The recent announcement regarding its planned Thor product with 2,000 TOPS of computing power is an order of magnitude more powerful than most of the chips on the market today aimed at ADAS applications.
The industry is already starting to market its models based on the performance of its autonomous features. Polestar, the electric branch of Swedish automaker Volvo, has recently added details about its front-sensing capabilities to its advertising campaign. It boasts that its front radar/camera combination has a range of 200m and a field of view of 45˚. This is fairly new, and it is hard to think of another car company advertising based on its sensing capabilities.
No doubt we’ll see more of this type of marketing from other automaker brands. The basic discussion to date has been around range and hence battery life. We’re not too far away from being able to compare in-cabin infotainment, and maybe even some of the OTA delivered additional services that software defined vehicles promise. But only if the consumer is prepared to pay extra over and above the price of the vehicle itself, that is.
You must or to post a comment.
This site uses Akismet to reduce spam. Learn how your comment data is processed.


Share this article


More from this Author