Jetson AGX Thor
An NVIDIA edge computing platform designed for robotics and automation scenarios. It features a next-generation GPU architecture and dedicated deep learning accelerators, delivering hundreds of TOPS of AI compute within a 60W power envelope, with high on-chip memory bandwidth and power efficiency.
Timeline (last 90 days)
On Jetson AGX Thor, d1-3B takes 16 milliseconds to answer one question and 20 milliseconds for three questions; on Jetson AGX Orin 64GB it is 26 milliseconds and 35 milliseconds; on Jetson Orin Nano it is 50 milliseconds and 73 milliseconds.
Edge-LLM based on TensorRT achieved a 6.4x speed improvement on the MLPerf edge Agentic benchmark on the Jetson AGX Thor.
The Jetson AGX Thor is equipped with NVIDIA's next-generation GPU architecture and a dedicated deep learning accelerator, with significant improvements in memory bandwidth and power efficiency compared to the previous-generation Jetson AGX Orin.
The Jetson AGX Thor is capable of delivering hundreds of TOPS of AI computing power within a power envelope not exceeding 60W.
All Facts (4)
On Jetson AGX Thor, d1-3B takes 16 milliseconds to answer one question and 20 milliseconds for three questions; on Jetson AGX Orin 64GB it is 26 milliseconds and 35 milliseconds; on Jetson Orin Nano it is 50 milliseconds and 73 milliseconds.
50%UnverifiedEdge-LLM based on TensorRT achieved a 6.4x speed improvement on the MLPerf edge Agentic benchmark on the Jetson AGX Thor.
50%UnverifiedThe Jetson AGX Thor is equipped with NVIDIA's next-generation GPU architecture and a dedicated deep learning accelerator, with significant improvements in memory bandwidth and power efficiency compared to the previous-generation Jetson AGX Orin.
50%UnverifiedThe Jetson AGX Thor is capable of delivering hundreds of TOPS of AI computing power within a power envelope not exceeding 60W.
50%