The goal was to deliver heightened performance to Edison users. The board itself has more power than the competition, but wouldn't it be better to have even more? The program is essentially a social big-data application, running python map-reduce jobs across any number of Edison boards. This would allow the unused cpu cycles and memory of hundreds of boards to be used together for large, complex calculations.

There are 3 parts:

On-board SW: suite of python scripts which listen for network signals from the server, and report back status or take on jobs

API: Node.js API for scheduling jobs, creating clusters, and accessing your device online

Front End: Discover other people nearby (or with low pings) and request their board in your cluster. Manage jobs and clusters from an easy to use web interface. Check it out at CrowdCompute.me

Share this project:

Updates