Hi,

I run SVD on a 500*1000 dense double matrix of random number, with
thin U and V, and it takes about 26 seconds. The code looks like:

JacobiSVD<MatrixXd,HouseholderQRPreconditioner> svd(In,
ComputeThinU | ComputeThinV);

On matlab it takes about 1.1 seconds, though admittedly the solution
doesn't seem terribly accurate.

What are possible reasons for this disparity, and is there anything I
can do to improve the situation?

Hugh

Are you running your code in Debug or Release mode? Debug mode can be 1-2 orders of magnitude slower.