The CGN acquires and manages its enormous constituent capacity of worker nodes by taking an innovative approach.
The CGN amasses and controls a vast pool of idle or sporadically available CPU power, opportunistically acquired from a wide range of available machines, which may be anything from a user’s home PC, the excess capacity of a data center.
While some of this hardware may be high-grade capacity that’s mostly completely available, that’s not a requirement for the excess capacity to be useful to the CGN.
The CGN will be able to work with compute hardware that may be inherently unpredictable, i.e. will make use of machines that cannot offer any guarantee of consistency of performance over time, and which may offer continually varying amounts of CPU power to the CGN, depending on other activities taking place on the computer.
The CGN dynamically restructures these individual resources that may vary in quality into a global network of quality compute power that is delivered as a well-managed, coherent service, presenting itself to Web applications as a solid, dependable services layer, one which is well-managed, is highly available, runs worldwide, offers high capacity, is scalable, and offers edge support