in GPU there are large number of processors as I know. How I can assign each iteration to each processor separately.

I need to give one kl value for one processor and another one to another processor

% my code
  kl= gpuArray(1:10000)
  lb=[rt(kl) 93500+10*kl ry(kl) 25e-6+2*rt(kl) ru(kl)];

    a0=lb;
    options=optimset('Display','iter','TolFun',10e-0,'TolX',10e-0,'Algorithm','trust-region-reflective');

    ff=@(a)myfunstdNBPpulsev1(a,e,dat);

    %ub =[20e-6 93900 10e7 75.1e-6 10.1e5]; %peak-1
    ub = [2e-6+rt(kl) 93600+10*kl 1e7+ry(kl) 27e-6+2*rt(kl) 1e5+ru(kl)];

    [a,residual]= lsqnonlin(ff,a0,lb,ub,options);