Skip to main content

Jetson AGX Thor Series

NVIDIA® Jetson Thor™ series modules give you the ultimate platform for physical AI and robotics, delivering up to 2070 FP4 TFLOPS of AI compute and 128 GB of memory with power configurable between 40 W and 130 W. They deliver over 7.5x higher AI compute than NVIDIA AGX Orin™, with 3.5x better energy efficiency.

NVIDIA Jetson AGX Thor module, front view

Technical specifications​

SpecificationJetson T5000Jetson T4000
AI performance2070 TFLOPS (FP4, sparse)1200 TFLOPS (FP4, sparse)
GPU2560-core NVIDIA Blackwell GPU with fifth-gen Tensor Cores, MIG with 10 TPCs1536-core NVIDIA Blackwell GPU with fifth-gen Tensor Cores, MIG with 6 TPCs
CPU14-core Arm Neoverse-V3AE 64-bit, 1 MB L2 per core + 16 MB shared L312-core Arm Neoverse-V3AE 64-bit, 1 MB L2 per core + 16 MB shared L3
Memory128 GB 256-bit LPDDR5X, 273 GB/s64 GB 256-bit LPDDR5X, 273 GB/s
StorageNVMe through PCIe, SSD through USB 3.2NVMe through PCIe, SSD through USB 3.2
Networking4x 25GbE3x 25GbE
Power40 W – 130 W40 W – 70 W
Connector699-pin B2B, integrated thermal transfer plate with heatpipe699-pin B2B, integrated thermal transfer plate with heatpipe
Dimensions100 mm x 87 mm100 mm x 87 mm

Source: NVIDIA – Jetson Thor