Nowadays, the thermo-energetic design of a machine tool also includes the thermal stabilization of its machine components. In the past, thermal stability was irrefutable connected to minimizing the temperature gradient of machine tools by air conditioning the entire machine or even the factory hall. Today, thermal stability also defines minimal inhomogeneities in the temperature field of the machine tool due to higher energy efficiency requirements. Fluidic tempering systems of machine components offer considerable potential concerning the minimization of thermo-elastic displacements with acceptable energy demand. Hence, intelligent algorithms are required to combine tolerable geometric deviations with minimal energy effort. The scope of this paper is the integration of a demand-oriented fluidic temperature control system into a machine bed. The resulting multi-input-multi-output (MIMO) systems and varying boundary conditions are challenges, which are addressed. Therefore the paper compares two control approaches, a decentralized single-loop control and a multi-loop control by decoupling the control loops and especially focus on the distribution of the jointly used actuating variable, combined with the variation of different boundary conditions.