Introducing the Jetson AGX Orin Series, and Jetson AGX Orin Developer Kit Availability

Autonomous Machines Jetson & Embedded Systems Jetson AGX Xavier
1

NVIDIA is pleased to announce that the new Jetson AGX Orin Developer Kit is available to order starting today*. Additionally we are announcing two Jetson AGX Orin series production modules available in Q4 2022.

  • Jetson AGX Orin 64GB with up to 275 TOPS AI performance in 15 to 60 W
  • Jetson AGX Orin 32GB with up to 200 TOPS AI performance in 15 to 40 W

The Jetson AGX Orin series is form factor and pin compatible with the Jetson AGX Xavier series. These modules can bring your Jetson AGX Xavier designs to the next level with up to 8X the performance, or up to 6X performance at the same price.

Summary of Jetson AGX Xavier and Jetson AGX Orin Comparison

JAXJAOForumTable
JAXJAOForumTable1366×469 93.2 KB

The Jetson AGX Orin Developer Kit includes a high-performance, power-efficient Jetson AGX Orin module, and can emulate all the Jetson Orin modules. The developer kit has the full GPU, CPU, DLA, NVENC, and NVDEC performance of Jetson AGX Orin 64GB, but with 32 GB memory.

See more details at our webpage: here, or additional documentation at the Jetson Download Center: here

* Please be advised there may be shipping delays due to short-term logistics issues.

1 Like
5

The powerful AI performance of AGX Orin is based on INT8, so I have to convert my model to INT8 . But I can’t convert MaskRCNN to INT8 either by torch–> int8 nor torch—>onnx–>tensorRT, would you like to share a successful demo with pytroch format? Thanks a lot.