The Neural Compute Stick is likely to have a range of applications.
The Movidius Neural Network Compute Stick is described by Intel as the world’s first USB-based deep learning inference kit and self-contained AI accelerator. You might remember that Intel acquired machine vision specialist Movidius Technology nearly a year ago.
Getting acquired from Intel has bestowed Movidius a few more flexibility in adding features such as the attainment to plug in a few of the compute sticks to accumulate more deep learning power.
“This enables a wide range of AI applications to be deployed offline“, he added, noting the benefit of having the ability to run systems on the edge independent to a central server or cloud.
With the new Movidius Neural Compute Stick developers can do things like train artificial neural networks on the Intel Nervana cloud, optimise existing workloads for AI as well as virtual and augmented reality and even automated driving.
If you are a deep learning developer you will still do the algorithm training on much more powerful equipment. The USB toolkit compiles and accelerates “neural networks at the edge”, the company said. While Intel itself is one of the Silicon Valley firms dominating AI research, it used the Movidius launch to stress it’s “committed” to giving smaller developers access to neural networks.
Intel has revealed the Movidius Neural Compute Stick, a USB-based kit created to help ease the development and deployment of artificial intelligence (AI) systems for developers and researchers.
The Movidius Neural Compute Stick is available to purchase as of writing priced at $79, down from the $99 Movidius originally planned to charge for its Fathom stick (Intel’s involvement clearly helped drive the price down a bit).
According to AnandTech, which seems to have more information on the device than the product’s own website, this year’s Movidius stick is slower than the original (150 GFLOPS vs. 100 GFLOPS), but has far more memory (4Gb vs. 1Gb).