The SSD market is very active because it combines a number of factors: new suppliers, VC funding, OEM deals, take-overs and rumours of acquisitions. In this note we look at the current activity levels and whether the market will settle down.
New established tier
SSD (Solid State Drive) is the semiconductor based alternative to HDD (rotating hard disks) connecting in the disk array or directly on the server. It is attractive since it is much faster than magnetic HDDs and prices are coming down. IT professionals exploit SSD by enhancing workload performance by placing hyper active files on an SSD tier. This is why all disk system vendors have introduced these SSD tiers on their disk systems. Most data centre IT professionals will by now have acquainted themselves with their current suppliers’ offerings. But there is still plenty of education to undertake regarding how to exploit SSD and not least understanding the potential issues with SSD.
United in criticising
It is interesting to observe how the SSD industry players are zig zagging between promoting the general endorsement of SSD and then pointing fingers at their competitors’ solutions. This is the sign of a relatively immature market, but also that there are some real issues with the technology. Those issues may well be fixed when the current semiconductor technology gets replaced. Until then, flash does not offer unlimited re-writes and the industry refers to a 2-3 year life cycle based on frequent access. There are plenty of use cases which can affect performance adversely which customers and the industry are busy discussing.
A number of vendors offer their own arrays: Violin Memory, Texas Memory Systems, Nimbus, Whiptail, SolidFire, Pure Storage. Recently EMC announced that it would acquire XtremIO. Then there are the OEM suppliers who sell their SSD drives to be embedded by the array vendors. There are also suppliers that produce PCI based SSD cards such as Fusion-io.
Many customers are adding an SSD tier to their storage pool in order to boost throughput and performance, typically when they upgrade to a new disk array. Eventually most data centres will have this SSD tier and this will apply to enterprises as well as SMEs. Some vendors have full size footprint products based on SSD, but only the most demanding workloads will justify those.
Scope for disruption
There is scope for vendor consolidation in the market which will have a calming effect, but there is also scope for disruption as rumours circulate about the replacement of flash. But flash has actually had longevity as it was invented as far back as in 1980 at Toshiba. So despite its flaws, coming up with an alternative that is price competitive has proven difficult. This means that customers should currently do their homework in understanding exactly how different the SSD offerings will support their workloads. And by the time today’s storage purchase gets faced out in 4-5 years’ time, we may well no longer be discussing flash.
Image credit: Manzzan