Two-dimensional magnetic recording is a novel technology proposed to extend the life of conventional granular magnetic recording (CGMR). It is the readback counterpart technology to shingled magnetic recording (SMR) that is being pursued by the industry today. In SMR, a 2-D block of bits is written as a single unit: modifying any bit within the block requires a modification to the entire 2-D block. In TDMR, by comparison, a 2-D block is read and processed as a single unit. Latency during writing of a 2-D block is not a serious problem as the data is buffered in the RAM. Latency during readback, however, results in a performance degradation to the user who has to wait before getting his file. If there is a single reader head, 2-D readback necessitates multiple revolutions of the disk leading to a significant latency, which may not be acceptable to the user and is one of the reasons the industry is hesitating for 2-D magnetic recording (TDMR). Recent events have seen companies working on the development of multiple (two or three) reader heads that read multiple tracks simultaneously with a single pass of the head. Such double/triple reader head arrays could be used to support TDMR, obviating the need for multiple revolutions of the media and reducing the read latency. An alternative use for the multireader heads would be in an SMR scenario, where only a single track is readback at a time. In this case, the extra readers could be used to estimate and mitigate the intertrack interference coming from the adjacent tracks. The difference between these two modes is the number of tracks detected with each head pass. For TDMR there would be two or three tracks detected with a single pass while for SMR there would be just one. In this paper, we study and optimize the impact of parameters of the double/triple reader head array being used in the SMR and TDMR readback scenarios. The key difference between these scenarios is in the design of the GPR equalizer and we compare and c- ntrast the recording system varying several key design parameters.