Silicon matters and not just switching ASICs.
Today Intel subsidiary Movidius is launching their Neural Compute Stick (NCS), a version of which was showcased earlier this year at CES 2017. The Movidius NCS adds to Intel’s deep learning and AI development portfolio, building off of Movidius’ April 2016 launch of the Fathom NCS and Intel’s later acquisition of Movidius itself in September 2016. As Intel states, the Movidius NCS is “the world’s first self-contained AI accelerator in a USB format,” and is designed to allow host devices to process deep neural networks natively – or in other words, at the edge. In turn, this provides developers and researchers with a low power and low cost method to develop and optimize various offline AI applications.
These devices could be used to improve the control and management plane software services.
- Network devices use 20-40 year old concepts using simple math for path determination and selection. If AI-like computation can be accelerated for $80 retail, what new services could the network offer for telemetry and path selection ?
- We already use ASICs for high speed path forwarding and FPGA for hardware control, why not add another chip for new services ?
- Power consumption is a concern for some people, dedicated hardware most often reduces power consumption.
- What about security functions such as IDS/IPS, malware etc where the detection and reaction loops could be much faster.
- Intel doesn’t always succeed with these devices, and
- most networking vendors struggle to have decent code with Java or C. Its doubtful they are prepared for AI code.
- Standards bodies are key for interoperability which would be needed here.
So, at best, an interesting thought exercise.
Intel Launches Movidius Neural Compute Stick: Deep Learning and AI on a $79 USB Stick : http://www.anandtech.com/show/11649/intel-launches-movidius-neural-compute-stick