Keeping in view the architecture of a disk drive, we know that not only disk spins under the head but also head of disk has to move forward and backward to access required track on disk. So the time that disk head will take to access the track is called seek time and this process is called disk seeks. When data is read sequentially the seek time is minimal and if it is being read randomly then seek time may be more. Seek time affects I/O performance of your database system. So to keep the seek time minimal it is important to disperse your data files on multiple disks so that sequential read is performed to maximum extend.
Click here to read all posts related to SQL Server storage systems and I/O operations