An estimated 5% of the memory market goes into automobiles today and this is expected to grow to 10% within the next few years as cars use more and more electronic functions to provide greater safety, greater efficiency, driver assist capabilities, communication between vehicles and local networks and richer telemetric and entertainment functions. Today about 40% of the value of an internal combustion engine automobile is in the electronics. For a hybrid or electric car this rises to about 75%.
Cars are becoming rolling application platforms and with the increase in the intelligence of on-board computers and connections between cars and local networks, they can participate in local communication and processing. This is often called fog or edge computing. Likewise they can be connected to the Internet and Cloud Computing services to provide vaster amounts of storage and processing power.
Cars are becoming big-data generators. To give an idea how this could develop, note that Google autonomous cars generate about 1 GB/sec. The average US driver drives about 600 hours/year. Thus a typical US driver of a autonomous vehicle could generate about 2 PB/car per year. With millions of cars on the road in the US the total amount of data generated is enormous. This data, or a sub-set of it, would most likely be stored and processed in the cloud.
Cars are an important element in what is often called the Internet of Things comprised of interconnected intelligent devices. The design of these systems to provide strong encryption and privacy protection can make ensure that our cars are our servants rather than someone else’s spies. At the same time, connected intelligent automobiles will enable many consumer services, such as streaming content, traffic alerts, road-side assistance and theft protection. They can also provide enterprise services such as warranty maintenance, fleet optimization and even interesting services such as usage-based insurance (pay as you drive).
All these applications for cars require memory and storage. This memory can be of various types. SRAM, DRAM and EEPROM are used for operating and code memory. Likewise NOR and SLC NAND flash memory can be used for code storage. NAND flash is used for operating storage and flash based SSDs and even HDDs are used for mass storage (particularly for infotainment systems).
Several companies participate in the automotive memory and storage market including Micron, SanDisk and Toshiba. Micron makes various volatile memories as well as NOR and NAND flash and SanDisk makes NAND flash memory. Toshiba makes a special HDD with up to 320 GB capacity, primarily for infotainment applications. There are also automotive applications using new memory technologies using MRAM products made by Everspin.
According to Micron the overall growth in the memory market is expected to grow by 12% annually. As shown below, Infotainment (information and entertainment services) memory needs are expected to grow by 12%, the instrument cluster memory needs are growing by 15% annually and Advanced Drive Assistance System (ADAS) memory needs are growing by 20% annually.
The Internet of Things will be a great driver for memory and storage technology both at the source device and in the network and supports and uses the data from that device. The promise of safe intelligent and autonomous transportation is impossible without significant memory and storage capabilities offered by today’s and tomorrow’s digital memory and storage technologies.
