The key is the k passes. This bounds the time from below for large values of k since it typically takes 10's of seconds to light up a map-reduce job. Larger clusters can actually be worse for this computation because of that.
On Wed, Apr 6, 2011 at 11:16 AM, Jake Mannix <jake.man...@gmail.com> wrote: > ... Lanczos-based SVD, for k singular > values, requires k passes over the data, and each row which has d non-zero > entries will do d^2 computations in each pass. ... > > I guess "how long" depends on how big the cluster is! >
