Star formation is intimately linked to the dynamical evolution of molecular clouds . Turbulent fragmentation determines where and when protostellar cores form , and how they contract and grow in mass via competitive accretion from the surrounding cloud material . This process is investigated , using numerical models of self-gravitating molecular cloud dynamics , where no turbulent support is included , where turbulence is allowed to decay freely , and where it is continuously replenished on large , intermediate and small scales , respectively . Molecular cloud regions without turbulent driving sources , or where turbulence is driven on large scales , exhibit rapid and efficient star formation in a clustered mode , whereas interstellar turbulence that carries most energy on small scales results in isolated star formation with low efficiency . The clump mass spectrum of shock-generated density fluctuations in pure hydrodynamic , supersonic turbulence is not well fit by a power law , and it is too steep at the high-mass end to be in agreement with the observational data . When gravity is included in the turbulence models , local collapse occurs , and the spectrum extends towards larger masses as clumps merge together , a power-law description dN / dM \propto M ^ { \nu } becomes possible with slope \nu \lesssim - 2 . In the case of pure gravitational contraction , i.e . in regions without turbulent support , the clump mass spectrum is shallower with \nu \approx - 3 / 2 . The mass spectrum of protostellar cores in regions without turbulent support and where turbulence is replenished on large-scales , however , is well described by a log-normal or by multiple power laws , similar to the stellar IMF at low and intermediate masses . The model clusters are not massive enough to allow for comparison with the high-mass part of the IMF . In the case of small-scale turbulence , the core mass spectrum is too flat compared to the IMF for all masses .