Noticias
Intel’s USB stick for developing deep neural networks
“As well as for use by data scientists and academics, the device is ideal for developers and engineers looking at integrating computer-vision and artificial intelligence capabilities into their IoT,” according to RS.
Like the earlier Movidius Neural Compute Stick, NCS 2 is based on the Intel’s Movidius vision processing unit – but this time the latest Movidius Myriad X, which has a dedicated hardware accelerator for deep neural network (DNN) inference.
Advertisement
SIGLENT SVA1000X series
NCS 2 is a prototyping and development tool when combined with the Intel’s distribution of OpenVINO toolkit, which offers support for deep learning, computer vision and hardware acceleration towards “human-like vision capabilities”, according to RS. “The combination of NCS 2 and OpenVINO toolkit enables development from prototyping trained DNNs on the compute stick to porting DNNs to an Movidius-based embedded device or system with minimal code changes.”
The stick also supports Caffe and TensorFlow software libraries.
Like the earlier Movidius Neural Compute Stick, NCS 2 is based on the Intel’s Movidius vision processing unit – but this time the latest Movidius Myriad X, which has a dedicated hardware accelerator for deep neural network (DNN) inference.
Advertisement
SIGLENT SVA1000X series
NCS 2 is a prototyping and development tool when combined with the Intel’s distribution of OpenVINO toolkit, which offers support for deep learning, computer vision and hardware acceleration towards “human-like vision capabilities”, according to RS. “The combination of NCS 2 and OpenVINO toolkit enables development from prototyping trained DNNs on the compute stick to porting DNNs to an Movidius-based embedded device or system with minimal code changes.”
The stick also supports Caffe and TensorFlow software libraries.