Proving the Business Case for the Internet of Things

Intel launches neural stick for prototyping vision at the edge

Steve Rogerson
November 20, 2018
At last week’s Intel AI DevCon artificial intelligence (AI) developer conference in Beijing, Intel introduced its Neural Compute Stick 2 (NCS 2) designed to build smarter AI algorithms and for prototyping computer vision at the network edge.
Based on the firm’s Movidius Myriad X vision processing unit (VPU) and supported by Intel’s distribution of the OpenVino toolkit, the NCS 2 can speed the development of deep neural networks inference applications while delivering a performance boost over the previous generation neural compute stick.
The stick enables deep neural network testing, tuning and prototyping, so developers can go from prototyping into production leveraging a range of Intel vision accelerator form factors in real-world applications.
“The first-generation Intel Neural Compute Stick sparked an entire community of AI developers into action with a form factor and price that didn’t exist before,” said Naveen Rao, Intel corporate vice president. “We’re excited to see what the community creates next with the strong enhancement to compute power enabled with the new Intel Neural Compute Stick 2.”
Bringing computer vision and AI to IoT and edge device prototypes is eased with the capabilities of the NCS 2. For developers working on a smart camera, drone, industrial robot or smart home device, the stick can provide what’s needed to prototype faster and smarter.
What looks like a standard USB thumb drive hides much more inside. It is powered by the latest generation of Intel VPU – the Intel Movidius Myriad X VPU. This is the first to feature a neural compute engine – a dedicated hardware neural network inference accelerator delivering additional performance. Combined with the OpenVino toolkit supporting more networks, the NCS 2 provides developers prototyping flexibility. Additionally, thanks to the Intel AI In Production ecosystem, developers can port their NCS 2 prototypes to other form factors and productise their designs.
With a laptop and the NCS 2, developers can have their AI and computer vision applications up and running in minutes. The stick runs on a standard USB 3.0 port and requires no additional hardware, enabling users to convert seamlessly and then deploy PC-trained models to a wide range of devices natively and without internet or cloud connectivity.
The first-generation Intel NCS, launched in July 2017, has fuelled a community of tens of thousands of developers, been featured in more than 700 developer videos and been used in dozens of research papers.
At last week’s Intel AI DevCon Beijing, more than 1000 AI developers, researchers and Intel customers and supporters gathered to collaborate on the advancement of AI and hear the latest updates on Intel’s AI portfolio of technologies, including:

  • Cascade Lake, a future Intel Xeon scalable processor that will introduce Intel Optane DC persistent memory and a set of AI features called DL Boost. This embedded AI accelerator is expected to speed deep learning inference workloads, with enhanced image recognition compared with current Xeon scalable processors. Cascade Lake is targeted to ship this year and ramp in 2019.
  • Intel's Vision accelerator design products targeted at AI inference and analytics performance on edge devices come in two forms: one that features an array of Intel Movidius VPUs and one built on the Intel Arria 10 FPGA. The accelerator products build on the OpenVino toolkit that provides developers with improved neural network performance on various Intel products and helps them unlock cost-effective, real-time image analysis and intelligence within their IoT devices.
  • Spring Crest is the Intel Nervana neural network processor (NNP) that will be available in the market in 2019. The Nervana NNP family leverages compute characteristics specific for AI deep learning, such as dense matrix multiplies and custom interconnects for parallelism.