Cold dark matter (or CDM) is the improvement of the big bang theory that contains the additional assumption that most of the matter in the Universe consists of material that cannot be observed by its electromagnetic radiation (dark) and whose constituent particles move slowly (cold). As of 2006, most cosmologists favor the cold dark matter theory as a description of how the universe went from a smooth initial state at early times (as shown by the cosmic microwave background radiation), to the lumpy distribution of galaxies and their clusters we see today — the large-scale structure of the universe. The theory was originally published in 1984 by United States physicists Joel R. Primack, George Blumenthal, and Sandra Moore Faber.
Primordial intermediate mass black holes (created in the big bang instead of by mass accretion) between 30 and 300,000 solar masses in galactic halos are consistent with observations of wide binaries as well as microlensing and galactic disk stability.
In the cold dark matter theory, structure grows hierarchically, with small objects collapsing first and merging in a continuous hierarchy to form more and more massive objects. In the hot dark matter paradigm, popular in the early eighties, structure does not form hierarchically (bottom-up), but rather forms by fragmentation (top-down), with the largest superclusters forming first in flat pancake-like sheets and subsequently fragmenting into smaller pieces like our galaxy the Milky Way. The predictions of hot dark matter strongly disagree with observations of large-scale structure, whereas the cold dark matter paradigm is in general agreement with the observations.
Three important discrepancies between the predictions of the cold dark matter paradigm and observations of galaxies and their clustering in space have arisen, however, creating a potential crisis for the whole picture.
The cuspy halo problem is that cold dark matter predicts that the density distribution of DM halos be much more peaked than what is observed in galaxies by investigating their rotation curve.
All of these problems have a number of proposed solutions. However, it remains unclear whether they represent a real crisis for the CDM paradigm, or an indication that the model needs further development.
The CDM theory makes no predictions about exactly what the cold dark matter particles are, and one large weakness in the cold dark matter theory is that it is unclear what the dark matter consists of. The candidates fall into three categories which are "humorously" named, as is common in physics.
WIMPs or Weakly Interacting Massive Particles assume that the dark matter is some sort of heavy unknown particle. Unfortunately, there is no known particle with the required properties. The search for these involves attempts at direct detection by highly sensitive detectors and attempts at production by particle accelerators.
It is also thought that CDM may be made of purely gravitating inflationary relics - so-called dark matter "X-particles"  such as Holeums.
^ Frampton, Paul H. (2010) "Looking for Intermediate-Mass Black Holes" Nuclear Physics B - Proceedings Supplements 200-202:176-8, doi:10.1016/j.nuclphysbps.2010.02.080
Bertone, Gianfranco (2010). Particle Dark Matter: Observations, Models and Searches. Cambridge University Press. pp. 762. ISBN 13: 9780521763684.