The internet of cars is paved with silicon
// php echo do_shortcode (‘[responsivevoice_button voice=”US English Male” buttontext=”Listen to Post”]’)?>
Technologists often overuse words like “disruptive” and “revolutionary,” but recent developments in automotive design and manufacturing are undoubtedly both disruptive and revolutionary. These changes represent the most significant upheaval in transportation design in the past 100 years, the greatest advancement since the earliest days of cars, when steam, gasoline and even battery-powered vehicles mingled on the same roads. dusty. The entire car is being overhauled, this time with electronics, to remake cars as supercomputers on wheels.
A BMW engineer lamented that his team spent decades designing cars with exceptional handling and performance, but the first thing buyers want to know is whether the vehicle fits their cell phone. You can understand the frustration of traditional automotive designers as electronics and software replace mechanical design as the criteria for customer selection.
It is no coincidence that Mercedes-Benz, Nissan, Volvo and others run advertisements highlighting cordless phone chargers, fully digital instrument panels, collision avoidance systems and 24-channel stereos in their vehicles. cars. No one mentions the horsepower, revs, drive, handling or even the “rich Corinthian leather”. Even Ford truck ads tout the number of 12V outlets and the built-in generator, not the truck’s cargo capacity or towing capacity. GVWR (total maximum weight) has given way to teraflops and MHz.
Electric motors, autonomous driving technology, infotainment electronics, subscription services, aftermarket wireless updates and emerging vehicle-to-all communication have exponentially increased the demand for chips and semiconductor software. Industry analysts now treat automotive electronics as a separate category, a sub-industry in itself, like IT or aerospace.
Where is all this heading? Connectivity now required to the internet and cloud data centers is shifting core competencies of automakers from mechanical design to software and silicon. The Car Internet will not run on gasoline or even electricity but on data. This data is collected, processed, communicated and stored by system-on-chip (SoC) semiconductors.
Automatic SoCs drive demand
Global automakers produce around 100 million cars and trucks each year. According to research firm IHS Markit (October 2020), more than half of them will include 15 to 24 complex electronic SoC devices by 2026, with an average of 23 SoCs per vehicle. Multiply those 23 SoCs by about 60 million cars, and you get almost 1.5 billion complex SoCs a year for the auto market alone, and that’s only five years from now. In the early 2030s, 10 years from now, demand will grow to over 2 billion complex automotive SoCs.
The Car Internet represents the most significant semiconductor opportunity since the smartphone. Some of this silicon goes into the advanced autonomous driving technology of Advanced Driver Assistance Systems (ADAS). Still, most go for infotainment and telematics, the most familiar and prevalent features of almost any car: radio, GPS navigation, electronic instrumentation, cell phone integration, engine management, and more. These are all must-have features in an automotive retail market driven by consumer demand, hence forward-thinking automotive advertising.
But this technology is only intended for the cars themselves, the “endpoints” in engineering jargon. What about the infrastructure behind them? Automakers need cavernous rooms full of computer servers, much like Amazon, Google, Facebook, or Microsoft.
If we assume one server for 200 cars, that equates to 300,000 servers per year. In addition, municipalities around the world are building infrastructure. Already, “smart city” technology is deployed in millions of traffic lights, road signs, cameras, street lights, road sensors and vehicle-infrastructure networks, supported by countless kilometers of wired access points and wireless. Cars will communicate with each other and with roadside base stations for latency sensitive feedback. All of this requires additional silicon.
Then there is the software. Conservative estimates place the onboard software stack at around 30 million lines of code. (For comparison, older versions of Microsoft Windows had around 50 million lines of code.) Some estimates are over 100 million lines for a new car with full functionality. Either way, it is a complex, networked, and highly reliable multiprocessor system.
The reliability of the equipment is crucial, as the automakers will be legally responsible for any dangerous failures. Chip designers respond with built-in safeguards and functional redundancy, in which an SoC has two or more internal processors running side by side, continually checking each other. This adds even more silicon and more complexity to the design of the SoC.
While commercially available chips are often sufficient, automakers can squeeze more performance from their hardware and add differentiation to their software if they design their own custom chips. Image processing neural networks, for example, need very specific, large and complex machine learning capabilities. An example of such a development is the latest Tesla Dojo SoC to turn millions of videos from the Tesla fleet into neural network learning to improve the user experience of automated driving.
Automotive decision systems work best with fully custom chips or Intellectual Property (IP) blocks embedded in semi-custom chips. This means that many of those 2 billion SoCs coming in the early 2030s will be fully or partially customized to meet the demands of the new automotive ecosystem.
Automotive companies around the world are looking to develop their own solutions, which is why many, if not most, end up working with companies such as ourselves (Arteris), Arm and CEVA, companies that license ownership. intellectual property for SoCs. Our vision for the direction of the automotive market is based on our experience of over a decade working with more than 30 customers in the automotive industry.
Data is king
It’s hard to say whether Tesla’s stock price is overvalued or undervalued, but there’s no question that the company exemplifies the electronics-driven approach to car manufacturing. It’s no coincidence that Tesla’s market cap is valued as if it were a Silicon Valley tech company. Following the lead of many consumer electronics companies, it designs many of its chips and develops its software. His main skill lies in electronics, not in metal bending. The company also manages its data center servers. Its cars are still connected to the internet and Tesla’s servers, with large amounts of data flowing back and forth in the form of maps, entertainment, traffic information, visual data collected by on-board sensors, and updates. software updates. This data is of immense value to Tesla, and other Internet of Car companies – all of them – will need to build similar systems.
Mobileye is another example of a company that uses data from its customers to constantly improve its technology based on cutting edge software, silicon SoC and machine learning techniques. Automakers don’t build everything from scratch, of course, but rely on well-established networks of suppliers and OEMs (Bosch, Continental, Denso, ZF, Magna and many others) who design and build major subsystems. These are mainly electromechanical devices, such as automatic transmissions or air conditioners. Going forward, these leading vendors will need to build SoC design and software development teams to continue delivering high value-added subsystems to automakers. Otherwise, other companies with more electronics experience will likely take their place. Many custom SoC developments will involve partnerships between traditional semiconductor design companies who know how to build SoCs and automotive OEMs and Tier 1 manufacturers who have application knowledge.
While the Car Internet will take time to mature, it has already emerged, powered not by electricity, hydrogen or gasoline but by data. Data drives the continuous improvement and evolution of the user experience, and consumers are willing to pay for this revolution. The core competencies of automotive manufacturers range from mechanical design of engines and chassis to software and silicon. The Car Internet is paved with silicon.
Charles Janac is President, President and CEO of Arteris IP is President and CEO of Arteris IP where he is responsible for growing and establishing a strong global presence for the company which is at the forefront of the NoC technology concept. Charlie’s career spans 20 years and spans many industries including electronics design automation, semiconductor equipment equipment, nanotechnology, industrial polymers, and venture capital.