It is good news for the developers now that the GPU manufacturing giant NVIDIA’s  “Jetson Orin NX 16 GB” is available for purchase in India that can be used to improve Edge AI performance. 

Designed for small robots, embedded devices, and autonomous machines like drones and handheld devices, it will be able to develop solutions using complex AI models in natural language understanding, 3D perception, and multi-sensor fusion.

The Jetson Orin NX is the smallest module of its kind and offers up to 100 TOPS of AI performance with the ability to adjust power usage between 10 W and 25 W. It’s also claimed to be three times faster than the NVIDIA Jetson AGX Xavier and five times faster than the NVIDIA Jetson Xavier NX.

The module includes features like NVIDIA’s Ampere architecture GPU, advanced deep learning and vision technology, high-speed input/output, and quick memory access to support multiple AI applications.

NVIDIA showed better results on computer vision tests using NVIDIA JetPack 5.1. The tests included running some dense INT8 and FP16 pre-trained models from NGC, and then comparing the results to the same models run on Jetson Xavier NX.

NVIDIA has announced improved results in computer vision tests using NVIDIA JetPack 5.1. The tests compared the performance of pre-trained models run on Jetson Xavier NX and Jetson Orin NX. The results showed that Jetson Orin NX had a 2.1 times performance increase compared to Jetson Xavier NX, with the potential to reach 3.1 times with future software optimizations. 

Additionally, Jetson Orin NX supports sparsity to enhance deep learning network performance. NVIDIA has also released JetPack 5.1 to support the Jetson Orin NX 16 GB and the latest CUDA-X stack.

NVIDIA announced the release of their unified quantum computing platform, “QODA” (Quantum Optimized Device Architecture), in July 2022 following the success of CUDA. QODA is intended to speed up advances in quantum research and development in AI, HPC, and other areas.