Automotive Power Concerns

By Ann Steffora Mutschler

With advanced semiconductor technologies infiltrating the automotive market in ever new and exciting ways, there are also challenges to implementation involving power. In fact, power has become a concern in many areas of automotive design.

Consider the Tesla, for example. The dashboard features a 17” touchscreen with the entire vehicle controls. This system is run by two Nvidia Tegra GPUs. Normally this would include a 16-bit CPU, but the CPU doesn’t offer enough performance.

The dashboard of the Tesla with 17" touchscreen display. (Source: Tesla Motors)

This isn’t such a simple tradeoff, though. The Tegras are designed for a tablet. As a result there are electromagnetic interference issues to consider. Moreover, tablets have two- and four-year lifecycles and they operate vastly different environments than an electric car dashboard.

“You have to take that same part and claim that it works from -50 to 150 degrees, that sort of range,” said Aveek Sarkar, vice president of product engineering and support at Apache “With a lot of these designs, taking the same design and applying it for multiple uses will create challenges.”

Sarkar pointed out that many semiconductor companies targeting the automotive market are starting to worry about power analysis and how to reduce the power consumption for chips aimed at automobiles.

Mladen Nizic, engineering director for mixed signal solution at Cadence noted that some automotive devices are power-sensitive by nature. “If you take a tire pressure monitor, for example, it has to collect the energy from motion and it has to be very power-efficient because there is no power supply inside the tire.”

And if there were a system that was monitoring the car continuously, it might actually drain the battery, so there are specific applications that are almost equally low power as a wireless phone, he said. “You have to try to minimize consumption as much as possible.”

Dollars and cents

Alexandre Palus, principal SoC architect at Altera, agreed. He said that in automotive, power translates directly to cash. “BMW and Mercedes have quantified the mileage per gallon that is equivalent to the watt. So depending on how many watts you consume, they will tell you how many miles per gallon you are going to get. At the end of the day, because the European car manufacturer has to pay a tax—the carbon tax—that translates to a certain amount of money from them to pay because of your device. So now when you are selling a device to a carmaker and you say, ‘I’m 2 or 3 or 4 watts,’ or whatever, and to them it’s the amount of money they are going to pay back so the [chip] that consumes less. It is actually a savings for them.”

Because European carmakers are doing this, it may only be a matter of time before it’s also standard practice with Japanese carmakers and in the United States, as well.

Related to power is EMI radiation, which is also a huge concern. “Depending on which application, for instance typically Bluetooth and Zigbee, these are already embedded. Broadband RF is likely next as there are higher speed radios, higher energy – the electromagnetic interference (EMI) is going to become a worse problem,” Sarkar said.

Thermal analysis—how heat effects the chip—is another area gaining traction among semiconductor vendors with devices aimed at the automotive space.

While these are just a few examples of the power challenges in the automotive space, it also represents new opportunities for EDA tools tailored to the specific requirements of this evolving and sophisticated market.