Will SSDs require efficient primary storage deduplication to bring them in line with hard disk pricing and help drive widespread adoption?
Typically, primary storage deduplication is performed either on the host server or at the storage controller level, not the individual drive level. I don’t believe that deduplication on SSDs will be required to drive increased adoption of SSDs. SSD prices have been decreasing and SSD capacities have been increasing over the last few years. I believe that with these factors, plus the significantly better performance and latencies of SSDs, adoption of SSDs will naturally increase.
HDD pricing has also generally been dropping, except for major supply disruptions, so it’s difficult to say that in general that SSD prices (per gigabyte) will be the same as HDD prices (per gigabyte) across the board in the near future. However, if primary storage deduplication were to be implemented at the drive level, then it could potentially increase the attractiveness of SSDs because it would have the potential to eliminate redundant writes, extend the life of flash media, and potentially allow for increased effective capacities to be advertised, depending on the deduplication algorithm and resulting ratio.
This was first published in January 2012