Not yet reviewed by Pith; the record is open.

The Via Lactea simulation of the dark matter halo of the Milky Way predicts the existence of many thousands of bound subhalos distributed approximately with equal mass per decade of mass. Here we show that: a) a similar steeply rising subhalo mass function is also present at redshift 0.5 in an elliptical-sized halo simulated with comparable resolution in a different cosmology. Compared to Via Lactea, this run produces nearly a factor of two more subhalos with large circular velocities; b) the fraction of Via Lactea mass brought in by subhalos that have a surviving bound remnant today with present-day peak circular velocity Vmax>2 km/s (>10 km/s) is 45% (30%); c) because of tidal mass loss, the number of subhalos surviving today that reached a peak circular velocity of >10 km/s throughout their lifetime exceeds half a thousand, five times larger than their present-day abundance and more than twenty times larger than the number of known satellites of the Milky Way; e) unless the circular velocity profiles of Galactic satellites peak at values significantly higher that expected from the stellar line-of-sight velocity dispersion, only about one in five subhalos with Vmax>20 km/s today must be housing a luminous dwarf; f) small dark matter clumps appear to be relatively inefficient at forming stars even well beyond the virial radius; g) the observed Milky Way satellites appear to follow the overall dark matter distribution of Via Lactea, while the largest simulated subhalos today are found preferentially at larger radii; h) subhalos have central densities that increase with Vmax and reach 0.1-0.3 Msun/pc3 comparable to the central densities inferred in dwarf spheroidals with core radii >250 pc.