There have been rumors that Tesla is developing its own chip optimized for self-driving applications ever since we reported on the company quietly hiring legendary chip architect Jim Keller from AMD last year.
Now further reports suggest that Tesla is indeed working on its own chip.
Sanjay Jha, CEO of GlobalFoundries, a semiconductor foundry, said at a conference today that they are working directly with Tesla on producing a chip.
Furthermore, CNBC reported today that Tesla is working with AMD on the chip:
“The carmaker has received back samples of the first implementation of its processor and is now running tests on it, said a source familiar with the matter.”
Update: Tesla and AMD refused to comment and GlobalFoundries denied the report.
That would make sense since Keller, who is now Tesla’s Vice-President of Autopilot Hardware Engineering, designed AMD’s latest chip architecture and the automaker poached a team of chip architects and executives from AMD following Keller’s hire.
So far, Tesla has been working with Nvidia on the computers powering the latest generation of its Autopilot program. As we recently reported, the automaker recently released a new Autopilot ‘2.5’ hardware suite with more computing power for autonomous driving.
Tesla claims that both its 2.0 and 2.5 suites could support fully autonomous driving, but it’s not stopping the company from working on even more powerful computing solutions to run its self-driving software.
The automaker’s Autopilot features, which they hope will eventually enable fully self-driving capabilities, are powered by Tesla’s own computer vision and neural net technologies, which are both demanding from a computing standpoint.
Those new reports follow other reports claiming that Tesla is making its own custom SoC (System on Chip) with Samsung.
Tesla didn’t respond to a request to comment on those reports.