A review study on the benefits of multi-reader heads in SMR and TDMR systems

Two Dimensional Magnetic Recording (TDMR) combined with Shingled Magnetic Recording (SMR) are upcoming recording technologies that are intended to see the industry over until more advanced recording schemes mature. SMR is a writing technology that enables writing of narrower tracks with wider write heads by means of shingling (overlapping) subsequent tracks. TDMR is the corresponding readback technology that enables reading of narrower tracks with wider read heads by using 2D signal processing to mitigate the inter-track interference that results. Despite these apparent similarities between them, TDMR would benefit much more a multiple-reader head than SMR would from a multiple writer which has motivated the production of multi-reader heads (MR heads) that until slightly more than a year ago, were thought to be impractical. The reason for this asymmetry is that during the writing process, data can be buffered in (volatile or non-volatile) memory and written in multiple-revolutions during SMR with no impact to the user. Reading multiple tracks over multiple revolutions on the other hand would produce tremendous latency overheads to the user that would render such a scheme unacceptable. Hence MR heads are currently being developed for eventual use in future HDD's. MR heads introduce several new parameters into the design of the HDD readback system that will impact on its performance. In this paper, several pieces of work pertaining to MR heads are reviewed that give us the framework to evaluate the expected performance of MR heads, a MR head architecture geometry to mitigate some potentially negative impact of skew on the system is suggested and a method to extract cross-track information when using multi-reader heads with an associated SMR or TDMR equalizer is proposed.