Solution to speed up loop with gamma distribution (gampdf)?

Dear readers,

Is there any solution to speed up this M-file if n = 1.000.000 ? Any help is much appreciate.

best regards,

Duy Nguyen

tic
% number of iteration
n = 1000;

% prepare parameters
a = 0; b = 100;
alpha = a+(b-a)*rand(n,1);
c = 0; d = 2000;
beta = c+(d-c)*rand(n,1);

% loop
t = 1:500;
TTD = zeros(numel(t),n);

for i = 1:n    
TTD(:,i) = gampdf(t,alpha(i),beta(i)); 
end
toc

% Elapsed time is 1.082072 seconds.

Answer 1

[Torsten]

100 sec to beat:

faktor = 1.0./(beta.^alpha.*gamma(alpha));
for i = 1:n 
  TTD(:,i) = faktor(i)*t.^(alpha(i)-1.0).*exp(-t/beta(i));
end

[le duy nguyen]

Although there are tiny differences between two TTDs, this solution is perfect to me. Thanks a lot!