New reports on Tesla self-driving car developments show that in a bid to upgrade the security and safety of these designs, Musk and company are going in-house to generate next-generation microprocessors.
Prior Tesla vehicles have utilized chips from Nvidia, a company with a good reputation in the microprocessor industry.
However, Stephen Shankland at CNet reports today that a new chip being made by Tesla will be 21 times faster than the previous Nvidia offering.
“Designing your own chips is hard,” Shankland writes. “But Tesla, one of the most aggressive developers of autonomous vehicle technology, thinks it’s worth it. The company shared details Tuesday about how it fine-tuned the design of its AI chips so two of them are smart enough to power its cars’ upcoming ‘full self-driving’ abilities.”
Engineers and others got a first look at these products in the Hot Chips Conference at Stanford yesterday.
Remarks from involved parties also show that the company understands the need to innovate in order to decrease risk in self-driving car designs.
There’s an intuitive understanding much repeated in the engineering world and in tech press that self-driving cars will need robust and redundant data processing in order to increase safety to the point where consumers and regulators will be comfortable letting these machines on the road.
Now, the development of proprietary in-house chips moves Tesla further toward that goal.
“There are a lot of redundancy features, which makes sure … nothing untoward happens to the system (in various failure scenarios)” says a chip designer, Ganesh Venkataramanan, according to Shankland’s story. “It was clear to us, in order to meet our performance levels at the power constraints and the form factor constraints we had, we had to design something of our own.”
In covering the process of development, Shankland further notes that Tesla’s ‘end to end’ acquisition of process development tools, from battery manufacturing to semiconductor engineering, gives that company a huge head start as new car designs drive industry competition.
With 36 trillion operations per second, each Tesla AI chip has 32 MB of high-speed SRAM memory.
Look for Tesla to continue to stay at the top of the class in terms of progress in the autonomous vehicle segment.