NVIDIA Launches Alpamayo 2 Super Open Reasoning Model for Robotaxis - NVIDIA Newsroom

NVIDIA unveiled its latest open reasoning model, Alpamayo 2 Super, at GTC Taipei, featuring 32 billion parameters and representing the most powerful open AI model for autonomous driving to date. This ... Read original →

NVIDIA’s Alpamayo 2 Super, unveiled at GTC Taiwan, China, isn’t just a parameter scale-up—it redefines the autonomous driving stack. Its 32B-parameter VLA architecture collapses perception, reasoning, and control into an end-to-end pipeline, directly undermining legacy modular software approaches and pressuring Tier 1s to adopt AI-native designs. While open-sourcing lowers entry barriers, it risks triggering renewed export controls in the U.S. and EU over high-level autonomy models, especially concerning RL training data and cross-border simulation flows. Against Waymo’s closed ecosystem and Tesla’s FSD v12, NVIDIA locks developers into DRIVE AGX Thor as the only viable deployment platform, cementing a hardware-software moat. Within 18 months, access to Alpamayo-class training-simulation-deployment pipelines will become a de facto gatekeeper for L4 robotaxi viability—excluding players lacking integration capacity.