LSQNONLIN with parallel programming

Hi, I am definitely new to parallel programming. I have a non linear set of algebraic equations (from 40 to 120 equations) which is created through a function. The ouput of such funtion is then a simple array of generally horrid equations. As it takes awful amount of times to reach convergence i am currently working with a very low precision. As far as I understood, I can use the parallel computing toolbox by adding:[ 'UseParallel' , true] in optimoptions, but waht I am not understanding is what else should I do. I am reading something about distributing arrays and gpuarrays... do I just call distributed arrays and gpuarrays functions for all the entries of LSQNONLIN? As you see I am pretty new to this... Can someone guide me to a tutorial or guide please? cheersLSQ