We show how data parallel operations can be implemented on a Global Cellular Automaton (GCA). For this purpose we implement a number of frequently used functions of the language APL. APL has been chosen because of the ability to express matrix and vector operations in a short and concise manner. It is shown how various operations on vectors and matrices can be executed through the execution of appropriate GCA generation computations. Finally the architecture of the underlying GCA machine is presented which interprets GCA instructions in a pipelined manner and using several memories.