The last few years have seen a massive drop off in the importance of traditional storage systems, culminating in the market leader (EMC) being acquired by Dell in 2015. Spending on hardware has dropped from $45b in 2011 to $33b in 2018 (see my Figure above). Shipments on the other hand have increased to reach their highest ever volume (907k) due to the introduction of cheaper arrays such as Dell EMC’s vxrack and vxrail and the growing use of storage software to improve the usability of installed arrays.
Measuring the revenues from hardware however is no longer the most important way of assessing vendor success. By adding software (which accounted for $12b in 2018) and hardware maintenance ($6b) we arrive at a total spend of $51b in the year (see my Figure above). Increasingly the functionality of storage used by businesses is being controlled through external software, sometimes addressing multi-vendor environments.
The solid state storage market is disaggregated with different players responsible for designing, patenting and licensing their innovations to a handful of major manufacturers, which makes it difficult to produce market share analyses. Nevertheless I’ve done my best to contrast and compare supplier business models to show the position in 2018. Here the market is subdivided by those producing disk drives, as well as NAND and DRAM solid state devices. Here there are high market entry barriers for new-comers due to the massive costs of building fabrication plants (currently around $4b) – with new ones having to be built for each new wafer size addressed. Consequentially each of the 3 sub-markets is dominated by a handful of vendors producing millions of chips or disk drives, which in turn creates stochasticism – with each new device type creating its own hype cycle of expectation, initial scarcity and high prices, followed by mass shipments and falling prices. As elsewhere in our industry, the fewer vendors there are, the greater the challenges of stochasticism are for users.
A significant proportion of raw storage devices find their way into storage systems arrays, although they are also widely used in client and mobile devices. In the last few years spending on disk drives has recovered, reaching $45b in 2018 – close to its peak in 2010 of $47b; despite this the disk drive market is far more stagnant than either NAND ($66b in 2018) or DRAM ($45b and due to overtake disk drive spending in 2019). The large profits being made by the largest silicon chip suppliers had been used for acquisition of many smaller vendors and the shake out of raw storage device suppliers hasn’t been reserved for the older disk drive market.
The most impressive part of the storage market is probably the incredible growth in capacity, which is driven on by constant (but expensive) improvements in the size of dies being manufactured. I show this growth of each of the 3 types of storage device in my Figure above. In total 2.4 zetabyes (ZB) were shipped in 2018. The underlying growth trend is exponential, although this is regularly interrupted when new products and/or new smaller die sizes reach production.
©ITCandor Limited – unauthorised copying of this content is illegal and will be rigorously defended by us through court action