Strip the keynote lighting and Nvidia’s automotive story has always been a platform story. The FY2020 10-K, filed this month and indexed by EdgarBeast, lays it out plainly at sec.gov: the company offers "DRIVE AV for autonomous driving and DRIVE IX for in-vehicle AI assistance," plus a scalable simulation product, NVIDIA DRIVE.
The interesting move here is the division of labor. DRIVE AV is the perception-and-planning brain; DRIVE IX is the in-cabin assistant. By naming them separately, Nvidia is telling you it intends to sell into both the safety-critical autonomy budget and the softer infotainment budget of a vehicle program — two different buyers inside the same automaker.
What the filing does not claim is just as instructive. There is no Nvidia robotaxi, no Nvidia-branded car, no promise to operate a fleet. The company positions itself one rung up the supply chain from the carmakers, which is the boring incumbent move in semiconductors: be the arms dealer, not the combatant. That posture is exactly what insulates it from the operational risk every AV operator carries.
The simulation product is the tell most readers skip. Autonomous-driving software needs billions of validation miles that are impractical to drive physically, so a simulator is not a nice-to-have — it is the only way to test edge cases at volume. Listing it alongside the runtime platform signals Nvidia understands the validation bottleneck is where AV programs actually stall.
Read forward from here and the thesis writes itself: a chipmaker selling the compute, the autonomy software, the in-cabin software, and the simulator to validate all of it. Whether automakers ultimately build their own stacks or rent Nvidia’s is the open question this 2020 filing sets up but cannot answer. The sec.gov text is the clean baseline; discovery via EdgarBeast.