I have to start by making sure my bias is out there, and it’s clear. I AM and Elon Musk die hard fan. He is the Visionary that allowed his dreams to become reality and in doing so disrupted several industries. I’ll leave that on the table, and give you the takeaways from the event on April 24, 2019.
Forget the bravado and consider what Musk says as a combination dreamer slash promoter which is okay. All the claims may have timing issues, but they set the bar for Tesla, as well as there competitors. The Autonomy Event was all about introducing the brand new microchips which are the best in the world and so on and so forth. Tesla announced that the new FSD computer (“full self-driving”) will be the catalyst for Tesla’s self-driving future. The claim was the microchips would be powerful enough to pilot the vehicles without supervision by the end of the year. Technically the chips can execute 144 trillion operations per second versus 7 million for Nvidia. Musk proclaimed that his prediction of having more than a million self-driving cars will be fully capable in a year from now.
The technical convergence of the conversation centers around a processor-centric model that doesn’t require 3D sensors (lidars). Most of Tesla’s competitors use lidars and assert they are essential in the implementation of self-driving vehicles.
From the comments I’ve seen, the chip isn’t perfect, and there are bugs in the implementation. There are instances where a car didn’t recognize a turn arrow on a traffic light and had issues at 4-way stops.
What is significant isn’t the new chip, rather the continued re-iteration of its goal of autonomy for vehicles in the foreseeable future. I’m not convinced it will happen in the next couple of years but it’s upon us.