Intel Neural Compute Stick 2 Launched for AI, Deep Learning Without the Cloud


Intel has released the next version of its Neural Compute Stick, a simple USB device that the company says can greatly accelerate AI processing and deep learning inference on consumer PCs. It is intended for use cases where neural networks need to be deployed without a connection to cloud-based computing resources. The new Neural Compute Stick 2 uses the Intel Movidius Myriad X VPU (visual processing unit). Intel acquired Movidius, a computer vision processing startup in 2016. The device costs $ 99 (approximately Rs. 7,137).

The Neural Compute Stick 2 is slightly larger than a thumb drive and is connected to a standard USB 3.0 port. The device has no fan and requires no additional power. It can be used with neural networks pre-trained for intelligent cameras, IoT devices, robotics, drones and RV hardware. It can also be used to test AI models "on the edge," which means on a local machine instead of in the cloud, for a fast return, or when sensitive data can not be sent to a remote server. Intel has already demonstrated demonstrations of the first-generation Neural Compute Stick by recognizing and classifying people and objects in a real-time video stream that can be used in security applications or for traffic management when low latency is important.

Intel claims up to 8X better performance in certain situations with Neural Compute Stick 2 compared to its predecessor. Myriad X VPU has a dedicated neural computing engine, 16 programmable computing cores and logic for processing 720p dual video streams. The device currently works with Linux on standard PCs or Raspberry Pi computers, but Intel said support for Windows ML is coming. The Tensorflow and Caffe frameworks are supported and Intel distributes its own version of the OpenVino toolkit to accelerate the performance of computer vision.

According to Intel, 43 percent of AI tasks will need to be processed on the edge, not in the cloud, by 2023, and there will be a 15X growth in the number of edge devices that support AI processing. The company expects artificial intelligence resources to become an important part of client PC workloads in the future. Almost 70 percent of Indian companies can deploy AI by 2020, according to the company's survey.

In addition to the acquisition of Movidius, Intel also bought the startup of AI Nervana Systems and the deep learning company Vertex.AI in recent years. Intel also recently announced that it has sold more than $ 1 billion in AI-capable processors in 2017. Recently the company held an AI DevCon conference in Bengaluru, where it announced that it has trained more than 99,000 Indian students, teachers and developers in AI and is working to improve the accessibility of IA development tools in India.

<! –



Source link