Its not a pipe. A pipe is used for two processes to communicate data, usually textual.
I'll give you a somewhat long answer so that you can understand it better. In linux, the "X window system", or X for short, forms the basis of the graphics system. The "X server" is responsible for ultimately drawing stuff on the screen. It interacts with your monitor, graphics card etc. Graphical programs are "X clients". They send data to the X server (which is a different program) and the X server displays it.
It turns out that X is "network transparent". This means that the client and server can be on different machines. The server, of course, must be on the same machine to which the display device is connected. So all you need to do is to tell your program which X server to connect to. You can do this by setting the DISPLAY environment variable
HOST is the name of the machine where the server is running. The 0 stands for "first screen" because there can be more than 1 X server running on the same machine.
Now start your program and the output will go to the remote machine.
To prevent people from gratuitously opening windows on other people's desktops (think of all the possibilities ;-), this is disabled by default. The user on whose screen you want to open a window must explicitly give you permission to do so by executing the command "xhost +CLIENT" where CLIENT is the name of the machine from which you want to run the program.