The curtains are up on “Tesla AI Day 2021”, and voila; there comes a person disguised as a humanoid robot dancing to the tunes of Elon Musk. The entire event was broadcast live for just over three hours. The first 50 minutes are spent with brilliant music, good sound effects, and the Tesla AV running on the road. The next 60 to 70 minutes explain the technology behind the Tesla FSD system, the Dojo supercomputer, the D1 chip, and the idea of a humanoid robot. Unfortunately, the deliverables were rocked amid the hype created around the Tesla AI event.
Chase its own tail
Two years ago in 2019, on Tesla Autonomy Day, Elon Musk, CEO of Tesla Motors, said that Tesla would have more than a million autonomous vehicles on the road where drivers “could go. ‘fall asleep’ during the trip and that too, by mid-2020. Hi Musk, it’s 2021 – we still have our hands on the wheels!
Image credits: You’re here
Forget that; now the American authorities are investigate approximately 765,000 Tesla Model X, Model S, Model Y and Model 3 vehicles to assess the technology and procedures used to monitor, assist and enforce driver engagement in the dynamic driving task when “autopilot” is used. Now Elon has presented a Tesla Bot named “Optimus”. Last time around we heard the term “Optimus Prime” in the famous Transformers movie series, where he is the leader of the Autobots. The name is probably adopted from the source. Take, for example, one of the most well-known humanoid robot companies – Boston Dynamics; it has not yet become fully productive and profitable for the industry.
Elon said, “Tesla bot will be real,” and the team hopes to release the prototype next year. The humanoid robot will be designed to perform repetitive tasks by taking advantage of Tesla’s existing technologies for automated vehicles. Yet another fanciful promise that could also miss the deadline – given past records.
Was it an ML conference?
Much of the event revolves around the technology behind Tesla self-driving cars. Of course, autonomous driving is not a new concept at all; However, explaining neural networks, including Neural Net Planner capable of handling the routing, trajectory and behavior of the car on the road, could be. In addition, Tesla engineers providing in-depth details on algorithms, sensors, multicam vector space predictions, RNN models, etc. are basically the hallmarks of an ML conference with ML engineers, data scientists as participants.
The focus of this part seems to give an impression of the robust safety features that Tesla vehicles take into account when making decisions during intelligent locomotion. Another aspect is to present the opportunities available at Tesla for ML engineers to work with an invitation – Join the Tesla AI team between events.
FSD systems and the technology behind them are present in all major automotive players, be it Waymo, Motional, Uber, Voyage or others. Tesla AI Day didn’t do anything new except for the fact that they brought behind-the-scenes work, mostly done in a closed environment in the foreground.
However, a key contribution is the announcement of the D1 chip for its Dojo supercomputer. Custom processors are based on 7nm processing technology with 50 billion transistors. The chip area is approximately 645mm², much smaller than that of NVIDIA A100 (826mm²) and AMD Arcturus (750mm²). It is equipped with 354 drive nodes and is powered by a quad-core 64-bit superscalar processor.
To sum up, all day Tesla AI could not show any substantial improvement over what we already have today.
Join our Discord server. Be part of an engaging online community. Join here.
Subscribe to our newsletter
Receive the latest updates and relevant offers by sharing your email.