In the present study, novel channel geometries in a wavy channel heat sink (HS) are investigated using ANSYS-FLUENT software. The Ag/water-ethylene glycol (50%) nanofluid is selected for cooling the CPU in this HS. The second-order upwind method is employed to discretize the momentum Equation and the SIMPLEC algorithm is employed for coupling velocity and pressure fields. Comparison of the two HSs with and without microtube shows that the presence of the microtube increases the uniformity of the CPU surface temperature distribution and decreases the mean surface temperature of the CPU (TCPU-Mean). However, the pumping power consumption of the system increases about 10 times. The results also demonstrate that the addition of nanoparticles results in intensification in the Performance Evaluation Criterion (PEC) of the system and up to 30%, especially at high Reynolds numbers.