Tesla Rebuilds Silicon Team With Leadership Hire for AI6 and Dojo 3 Chips

Tesla has rehired Anant Nivarti to lead its silicon engineering efforts, including development of AI6 and Dojo 3 chips. The move signals a revival of the company's in-house semiconductor program following the effective disbandment of its previous Dojo team. Nivarti's return marks a strategic shift for Tesla as it pursues custom chip designs for artificial intelligence and computing workloads. The Dojo 3 project specifically targets space applications, according to sources familiar with the effort. Tesla's Dojo project was initially developed as an in-house supercomputer for training artificial intelligence models used in the company's autonomous driving systems. The company unveiled Dojo in 2021 as part of its effort to reduce dependence on Nvidia graphics processors for AI training. The previous Dojo team was disbanded as Tesla adjusted its compute strategy amid broader cost-cutting measures. The revival under Nivarti suggests renewed commitment to custom silicon development as competition in electric vehicles and autonomous driving intensifies. Custom chip development allows Tesla to optimize hardware for its specific workloads, potentially improving performance and reducing costs compared to off-the-shelf alternatives. The AI6 chip likely represents the next generation of Tesla's in-vehicle AI processors, following the Hardware 4.0 and Full Self-Driving chips currently deployed in vehicles. Space applications for Dojo 3 suggest potential crossover with SpaceX operations, another company led by Tesla CEO Elon Musk. Both companies have previously shared engineering resources and technology development. The hire comes as semiconductor leadership changes accelerate across the technology industry, with companies investing heavily in custom silicon to differentiate their products.