NVIDIA has introduced the Jetson T3000 and T2000 modules, new edge AI computers based on the Thor architecture that bring powerful AI compute to mass-market robotics and autonomous machines.
The Jetson T3000 delivers 865 FP4 teraflops of AI performance in a compact form factor roughly half the size and power of the T5000 flagship module. It combines a Blackwell GPU with an eight-core Neoverse Arm CPU, 32GB of LPDDR5X memory, and 273GB/s of memory bandwidth. Despite its smaller footprint, the T3000 achieves similar inference performance for multimodal workloads including large language models and vision-language-action models.
The Jetson T2000 offers an entry point with 400 FP4 teraflops and 16GB of memory, targeting visual AI agents, autonomous mobile robots, and industrial manipulators. The new lineup gives NVIDIA a scalable edge AI platform spanning from 70 TOPS to 2,000 teraflops.
Leading robotics companies including 1X, Agile Robots, Amazon Robotics, Boston Dynamics, FANUC, and Hitachi are already building on the Jetson Thor platform. NVIDIA also released new Jetson agent skills that automate memory optimization across the entire portfolio, potentially saving weeks of manual tuning work.
The IGX T3000 variant delivers the same performance with integrated functional safety features and runs the NVIDIA Halos safety system for robots operating alongside humans, targeting industrial and medical applications.
