Movidius and Intel have put deep-learning on a stick with a tiny $79 USB device that makes bringing AI to hardware a snap.
In April of last year, Movidius showed off the first iteration of this device, which they then called the Fathom Neural Compute Stick. The company wasn’t able to get the product out as quickly as they had hoped because they were a little busy getting acquired by Intel.
Movidius’s goal has long been to move this type of image-based deep learning from the cloud to the edge with its Myriad 2 visual processing unit. The chips are used on everything from security cameras and drones to AR headsets, enabling them all to recognize and identify objects in the world around them.
The Movidius Neural Computer Stick tosses one of these VPUs into a USB 3.0 stick giving product developers and researchers the ability to enable prototyping, validation and deployment of inference applications offline, bringing about a number of latency and power consumption improvements. It supports
When connected to constrained host computing devices like Raspberry Pi, the compute stick offers pull-and-play intelligence. Interestingly this news comes just a month after Intel reportedly canceled plans to continue working on its own Edison, Joule and Galileo compute modules.
Getting acquired has offered Movidius some more flexibility to add features like the ability to plug in several of these sticks to add more deep-learning power. On a manufacturing side, Movidius has also been able to bring down the price from $99 to $79 even as its build has gotten more hefty.