Tesla launch chip to train AI models in data centers

Tesla to launch its in-house chip for the training of AI models in data centers

Tesla launch chip to train AI: Recently Tesla director Ganesh Venkataraman unveiled an innovation of Tesla’s and that innovation is a computer chip which is designed and built in the entire house, and the company is utilizing it for running their supercomputers which are known as Dojo.

The major part of Tesla’s Architecture Intelligent architecture is dependent on Dojo and the neural network training computer will be going to have the ability to process the large number of cameras imaging the data four times faster than other computing systems and this is stated by Elon Musk.

This ideology is the bring out the Dojo-trained Artificial Intelligence software will going to push out Tesla customers with the help of over-the-air updates.

Tesla launch chip to train AI, The chip which is recently unveiled by Tesla is known as D1 and contains the 7 nm technology. Ganesh Venkataraman has proudly held up the chip and stated that the GPU level is computing with the CPU connectivity and the twice of the I/O bandwidth which is the state-of-the-art networking switch chips that are used in today’s technology and are projected to be the gold standards.

He stated that he walked through the newly designed technical chip and also explaining that Tesla in the future wanted to own its stack in the tech field if possible by avoiding any bottlenecks.

Tesla has launched a next-gen computer chip last year which is developed and designed by Tesla and produced and manufactured by Samsung but somehow it has not been able to fulfill the shortage of chips globally faced by the auto industry for months.

Tesla launch chip to train AI, To survive this shortage Musk has stated that during an earnings call in the last summer the company has been forced to rewrite some of the automobile vehicles’ software after getting a substitute in alternative chips.

Rather than fulfilling the need for availability, the overall goal is to take the chip production in-house is to decrease the issue and build an Artificial Intelligence with better performance along with increasing its bandwidth. The reason is to develop a better supercomputing system Dojo more efficient.

Read More: Elon Musk will hopefully send Tesla Bots to work on Mars for the creation of inhospitable environments

About Bhavika

Sr. Consultant at Yugasa Software Labs - App | Web | IoT | AI
This entry was posted in News and tagged , , , , , . Bookmark the permalink.