With all the recent rumors about Tesla using different chips for its computing solutions to power the Autopilot and infotainment systems in its vehicle recently, there’s some confusion about Tesla’s current onboard computers.
Today, we clear things up with a new exclusive first look at the computing platform inside the Model 3.
When Tesla introduced its Autopilot 2.0 hardware suite last year, it was equipped with a powerful computer powered by Nvidia chips.
It was likely the most powerful computer in a production car at the time, but Tesla already updated the system with Autopilot 2.5 hardware suite introduced last month.
But as Tesla’s cars have increasing computer power needs, the company is continuously looking to improve on those systems. Latest rumors suggested they are looking to design their own chips for Autopilot and moving to Intel for the infotainment system.
Now we are able to confirm the exact computing solution that Tesla developed for Model 3 based on information from a reliable source with knowledge of the Model 3 program (the same who brought us our exclusive first look at Tesla Model 3 battery pack architecture).
Tesla developed a fairly impressive custom liquid-cooled dual computing platform with both its Autopilot and infotainment computers built onto two different boards in the same module:
On one side, there’s the infotainment Electronic Control Unit (ECU) or MCU now powered by the Intel Gordon Peak BMP and Intel Apollo Lake system-on-chip instead of Nvidia’s Tegra chip in Model S and Model X. And on the other side, there’s the Autopilot ECU still powered by the Nvidia GPGPUs.
It confirms the rumor that Tesla is switching to Intel to power its MCU and we now know the exact system used in the Model 3 for the first time.
Computing power inside cars is going to be an increasingly important feature that was long overlooked in the auto industry.
A significant amount of computing power is required to power autonomous driving and active safety features. Furthermore, with the advent of those features, the in-vehicle experience will become more important and therefore, more computing power will also be required to power that experience.
Like Apple unveiling more computing power in its iPhones or other consumer electronics with every iteration each year, I think we are going to start to see automakers doing the same with their vehicles.
I think Tesla bringing 3 different versions of its onboard computers to production over the last year is one of the early examples of that.
What do you think? Let us know in the comment section below.
Subscribe to Electrek on YouTube for exclusive videos and subscribe to the podcast.