Complete immersion of computer servers in dielectric coolant has become a promising solution to the Net zero initiative. However, insufficient data on thermal shadowing and thermal swapping effects on the computer server performance makes data center operators hesitant to fully accept and apply this technology. Thermal shadowing effects reflect the thermal field impacted mutually among components by the various configurations in the same server chassis; Thermal swapping effects take account of the mutual impacts among servers submerged in the same tank. CFD simulation has been proven as an effective approach to predict the performance of coupled heat and flow fields in the single phase, incompressible and Newtonian fluid. After being calibrated by the testing results, three cases abstracted from the typical deployments have been studied numerically, which are the original Dell PowerEdge R720 server in the single tank as a benchmark (Case1), Dell PowerEdge R720 with two NVIDIA Tesla K40M GPUs in the single tank (Case2) and a Dell PowerEdge R720 abreast of free space as its size (Case3). By comparison with the Junction temperature under the variation of coolant outlet temperature, the quantitative descriptions of thermal shadowing and thermal swapping respectively reveal that the computer servers designed for air cooling are not applicable or not optimal for the immersion cooling deployment, especially with the customized configurations. The add-on components will change the thermal field by different flow boundaries significantly. The preliminary lab test or simulation of servers must be undertaken. Different servers or the same server models with different configurations need to be categorized into different tanks referring to different flow-resistance characteristics. Under the mode of service, the existence of vacancy occupied by the server in service will worsen the thermal conditions of servers left in the tank. Generally, to promote immersion cooling technology flow-resistance curve under immersion conditions is recommended to be disclosed by server OEMs, and immersion cooling equipment suppliers/contracts must disclose the envelope of coolant outlet velocity and temperature under normal running and service modes.