SIMD-oriented Fast Mersenne Twister (or any other fast rng) for Monte Carlo?

Is there a faster PRNG for MATLAB that is suitable for Monte Carlo simulations, besides the Mersenne twister? For example, a MEX implementation of the SFMT (successor to the Mersenne twister) (as found on http://www.math.sci.hiroshima-u.ac.jp/~m-mat/MT/SFMT/)?

I am finding ways to speed up my Monte Carlo simulations, which are making a huge number of nested calls to randn.