Each year, around this time, experts in various tech sectors make predictions about what might occur in the coming year. Data storage experts, too, are happy to jump in with their prognostications. Those of you who are regular readers of my blog know that I often take a slightly different view of things – so, rather than adding to the long list of speculators telling us about what will happen in the coming year, I thought I’d tell you about 5 things that won’t be happening to data storage during 2016:
1) HDD-only storage arrays won’t survive
66 million years ago, an asteroid the size of Manhattan slammed into the Earth off the coast of Mexico, hastening the end of the 165 million-year-reign of dinosaurs. Similarly, in 2016, the effect of low cost cloud storage and high performance SSD storage systems will be felt around the globe, bringing about the demise of on premise mainframe-era (HDD-only) storage arrays. With costs falling and reliability rising, cloud and SSD storage systems are the 2016 equivalent of a new ecosystem evolving from the old lumbering storage dinosaurs of the past.
2) SSDs won’t all be flash
For roughly 50,000 years, humans got around slowly on foot, either their own, or on the feet of animals they domesticated. Then, two hundred years ago, new forms of propulsion were invented: steam locomotion, internal combustion engines, jet engines, and rocket propulsion. Human travel progressively accelerated as these new forms of transport were made available. An analogy could be made between the movement of people and the movement of data. From punch cards to paper tape, magnetic tape, magnetic disk, and solid state devices, the acceleration of data transport over the past 50 years has been profound. And just as people-moving vehicles are evolving towards engines powered by hydrogen and electricity, in 2016 we’ll see SSD engines powered by such technologies as RRAM, 3D XPoint, ST-MRAM.
3) The mighty little disk drive won’t go away
Referring back to the analogy in #2, you might have noticed that many people around the world are still mostly walking and/or riding on domesticated animals. Likewise, if you believe that SSDs are on the verge replacing all the HDDs in the world, here’s a reality check: during the 3rd calendar quarter of 2015, around 119 million HDDs shipped, while SSD shipments accounted for roughly 26 million units, of which a mere 3.1 million were of enterprise-class. SSDs are gaining on HDDs to be sure, but they still have a lot of catching up to do. For cold storage, cloud storage, and PC storage, it’s still hard to beat the economics of HDDs. For 2016, you can surely expect to see HDDs cross the 10TB threshold, and SSDs in the 4TB range. Any HDDs found in enterprise storage arrays will be supplemented by generous cache and/or SSD (see #1). Although SSDs will continue to find their way into high performance applications, by the end of 2016, they will still carry a cost 3-5 times higher than their HDD counterparts (on a raw cost-per-TB basis).
4) The Great and Powerful Cloud won’t go away
In the movie The Blob, a corrosive alien amoeba crashes from outer space and eats everything (and everyone) in its path, growing in size every time it consumes something. Sometimes things become so big that they seem to take on a life of their own. During 2015, the public cloud achieved this point of critical mass, reaching the size where it cannot be destroyed. In 2016, this means that enterprises will perceive less risk in moving data to the cloud, and we’ll see more applications running in the cloud. The business benefits of cloud are just too great to ignore: faster application ramp time, lower cost of failure, and overall increased competitiveness.
5) A multitude of clouds won’t work without a Data Fabric
Today’s cloud environment is a little like the story of the Tower of Babel – where a multitude of people building a great tower were confounded by not being able to understand each other, and eventually abandoned their project. Public clouds are today’s version of the Tower of Babel; with a multitude of clouds and no clear communication between them. To realize the full potential of the cloud, there must be a way to easily manage and move data between multiple clouds. The cloud providers themselves are not rushing to the assistance of people that want to move data from their cloud to someone else’s, so the solution will have to come from elsewhere. Will 2016 be the year that IT realizes the need for a data fabric? All signs point to “Yes”.
Opposing views are always welcome – feel free to leave a comment with your predictions (or non-predictions!)