The Grid++ Concept

The concept of Grid++ is very simple and yet very complex. Many computers attach to a Master Node (Grid++) and open two streaming channels: a data channel and a flops channel. The Grid++ services an algorithm that unifies these streams into one single stream, which we call the “Main Process”. The Grid++ analyzes the main process structure and splits the content of the main process into “Threads”. These threads are sent to at least 3 physical processors (attached computers). The first processor which returns the result is flagged to a higher rank. After a while the Grid++ has a clear “Computation Map” of which processor in which location should be used for what kind of thread. Now imagine a 100 processors computing all together in one big binary stream (also called parallel computing), the stream grows as more process nodes come alive and this continues as more computers attach them self to the Grid++. An amazing computation power is the result which we call the “Virtual Super Computer”. End users who are running Live Applications on a Grid System have basically almost unlimited execution power. This is basically an overview how Grid++ works, but there is a lot more involved like security, timing, database, etc. and everything has to be mapped into one big Computing Stream.

I hope this was informative.

If you want to know more about Grid++ and (X)Power++ then go to http://gridplusplus.com