Should chips like the Jetson Orin module be regarded as dual-use technologies? Or have similarly powerful embedded processors now become widely enough available that it isn't worth making the distinction any more?
The United States restricts the export of advanced NVIDIA artificial intelligence chips and high-performance data center GPUs to Russia