Tesla described Keller’s job as overseeing low voltage hardware. This is what Keller was good at, making the hardware platform that can make the company stand out from its competition. Elon Musk understood the importance of having its platform in the same way that Apple had with its iPhone by keeping software and most of the hardware development in-house. Fudzilla earlier reported that our sources are confident that Keller went to Intel and that will be announced shortly. 

Pete Bannon gets lead self-driving hardware job

The new guy who takes over Keller’s position is Pete Bannon. He is an ex Compaq, ex-Intel, ex-PA Semi and ex-Apple executive who worked very closely with Keller.

Tesla fans should not fear that Keller’s departure might cause a delay in the Tesla own self-driving platform. Keller probably did most of the job he was hired for and made a long-term roadmap. We are confident that people like Bannon can do a good job in the execution. Bannon has been with Tesla since February 2016, not long after Keller joined. He joined after close to eight years at Apple.

It is puzzling how Jensen at Nvidia is overseeing these staffing fluctuations. Musk wants to ditch Nvidia for its platform. Tesla's self-driving on the model three is being sold, but the technology does not work. You pay for something that is yet to happen, which seams to be a pattern in Tesla’s business model. After a few quarters of over-optimistic launch dates, Tesla has developed quite good products but tends to charge a fortune. The Tesla 3 $35.000 is a 49,000 USD car, but this is another story.

Tesla spokesperson told Electrek:

“Pete Bannon, who has been at Tesla for over two years, will now lead Autopilot hardware. Pete has been building processors since 1984, co-led the development of Apple’s A5 chip and then continued development through to the A9 chip. Prior to Apple, Pete was the VP of architecture and verification at PA Semi. Andrej Karpathy, Tesla’s Director of AI and Autopilot Vision, will now have overall responsibility for all Autopilot software.”

Andrey Kapathy runs the software

Andrey Karpathy continues to lead the computer vision and deep learning. His title as a Director of AI and Autopilot vision gets him to report directly to Elon Musk.

He is one of the star Ph.D. chaps from Stanford who worked on the “Convolutional Neural Networks for Visual Recognition.,” These boffins are in high demand and some of them making a lot of money.

We do know that despite the fact that Tesla gets most of the PR when it comes to self-driving, almost everyone else including GM is working on its own software solution. Most of them, including UBER and Co, plan to use existing hardware, such as Nvidia Drive solution and develop its softwre on top of that, while Elon’s team wants to make the whole thing, hardware and self-driving software in-house.