Memory Trends and Requirements in Connected and Autonomous Cars

As cars become more like computers, their need for faster processes will also increase. Memory is another key component, with infotainment needs rising from 16GB a few years ago to as many as 256GB today. That doesn’t even include connected or autonomous features, but estimates for the latter are quite significant.

“There are a lot of things that are driving the need for more storage within the vehicle,” said Scott Beekman, director of managed flash memory for Toshiba Memory America. “We estimate an autonomous vehicle may need between 1.5 to 3 terabytes of storage.”

Beekman noted that driverless cars are currently expected to generate as much as 4 terabytes per day from all the cameras, LiDAR, sonar and radar.

“All the functionality, whether it’s the infotainment, ADAS, even digital clusters, telematics – all this is needing more and more memory,” he said. “And then it’s not just more memory within the vehicle but then in the data centers as well. You need to record event data, how vehicles are [performing], and learn from that [through] data analysis, maybe improve the software.”

Automakers currently ditch a high percentage of vehicle data, and they are unlikely to continuously store a day’s worth of information – let alone a full month or year. Beekman thinks that would be a “very costly” investment.

Consumers have shown a willingness to pay for more storage, however. Smartphones are the perfect example with more than a billion units sold globally every year. This has driven the economies of scale for flash storage, allowing other applications to take advantage of the memory, Beekman explained.

Smartphones have also ushered in a new era of in-car functionality, inspiring OEMs to install larger touch screens with more features. In an age where cars drive themselves, smartphones and infotainment experiences are likely to converge.

With so much love for our pocket devices, automakers might be tempted to price their connected or autonomous cars in a similar manner. Where leather seats and sunroofs were once a differentiator, memory – and perhaps the speed of connectivity – could be key elements future cars. Even so, Beekman doesn’t expect that to have as much of an impact on cars as it has on phones.

“I think it will probably be a different model,” he explained. “When a consumer goes in and buys a smartphone, they can innately see, very quickly, that one is 64 gigabytes, the other is 128 gigabytes. They instantly recognize the value of that. ‘Ah, I can store more videos.’ Or they just want to be on the safe side and make sure they have more capacity, so the value is very obvious. I think when you buy a vehicle it’s not as obvious. You’re buying the vehicle and the flash is doing good things in there, but it’s in the background.”

Regardless of the pricing strategy, OEMs will have to evaluate the benefits that additional memory could provide. They must then attempt to convey those benefits to consumers, whether the buyer is an operator of a large automated fleet or an individual driver, who will ultimately decide what works.

“Is there an ROI in the consumers’ eyes for what’s being supported?” Beekman questioned. “I think clearly there’s huge value to the idea of an autonomous vehicle just in terms of increased productivity, safety and so forth. Memory is a big part of that.”