First - it’s a massive market. Cathie Wood of ARK Invest, suggests autonomous taxi platforms will generate $8 to $10 Trillion in 2030! Powering this drivetrain is a massive opportunity.
Second - it’s complicated. The twists and turns of the real world can make for life and death consequences. Training reliable AI models requires:
And the race is on, as tech companies both large and small, are in overdrive to solve these challenges. Here’s recent news about two of the big players.
Telsa isn’t just an electric car company; it’s an autonomous car company. An recent rumblings suggest it wants to be the “Amazon of Autonomy,” via sharing its self-driving tech platform to other vehicle manufacturers.
But here’s the twist: Tesla has been relying on NVIDIA to fuel their AI needs. That may change as Tesla develops its own “D1 Dojo” supercomputer chip. First announced in 2021, it went into production in July 2023, and recently doubled in production. The chip will accelerate Tesla’s machine-learning model with tons of real-life video captured by their vehicles.
AMD has always been in the hardware race, but has been outpaced by NVIDIA, whose overclocked software makes their chips easier to build, tune, and deploy.
But here’s the exciting news: AMD just hit turbo by acquiring Nod AI. Nod’s software stack is like reprograming the ECU on your existing ride. And not only will this speed up their autonomous division, Nod’s tech promises to rev up AMD’s gaming and graphics ability as well.
