We believe in flash. We do not believe in all flash storage arrays.
Parts of the post “Why All Flash Storage Arrays Have No Long-Term Future” were taken from an interview with our CEO, Boyan Ivanov, published in The Register on 19 Feb 2016.
Why all flash storage arrays have no long-term future? Data is expected to multiply 10 fold between 2013 and 2020 [1] with a CAGR of 26.4% through 2018, or about six times the growth rate of the overall IT market [2] Storage revenues of big players are decreasing [3] EMC, the largest storage company in the world, a market leader, and the #1 acquirer of storage technologies for the past years became an acquisition target itself. Disruption. The storage industry is turning on its head.
Storage customers realize that the old way of doing things is not sustainable anymore.
They do not want to pay crazy amount of dollars for something, which is becoming a commodity. They do not want to be locked-in. They want seamless scalability. They want simplicity.
They care about their data. They do not care about someone’s box.
Customers do not get these things with proprietary all-flash arrays. This is why the trend towards “software-defined storage” took the industry by storm. Or call it however you like – “Server SAN”, “software-enable storage”, “software driven/lead storage”. Does not matter. It’s the paradigm that shifts and analysts and industry pundits try to dress it up in a term they coined. Regardless, it’s the changing nature of business demands, use cases and customers’ needs that drives the market.
And we do not need another storage product – we need a new concept to storage altogether. And then this new concept is delivered as new products. We do not need another proprietary storage box. Be it “all-flash” or “all-SSD”. It’s still vendor lock in. It’s still expensive. It’s still single purpose, specialized piece of hardware. You still have to migrate data from an old box to a new box.
But wait: it has only SSDs inside! Big deal. Now every box can be an “all-SSD box” – any Dell/HP/IBM (well, Lenovo) server. Any Supermicro and Quanta, any white-box. Not a differentiator anymore. Sure, some of the “all-flash” products have clever software in the box, but some don’t. And it’s the software that makes the magic. All-flash arrays are just a short-term flick, part of the transition. Storage arrays in all forms will become the next mainframes – niche and rare. And for what will be the drives inside – HDDs, NAND flash, memristors – it’s not as important. The latest type of media will be adopted, as any new technology, every year. But the master trend is that there is a change in the approach to building data storage and this will last decades.
It’s all going to be storage software running on standard hardware, eventually. And “eventually” is happening now and it’s happening faster than you might think. Because people want something new. Something for themselves.
[1] http://www.emc.com/about/news/press/2014/20140409-01.htm[2] https://www.idc.com/prodserv/4Pillars/bigdata
[3] http://www.storagenewsletter.com/rubriques/market-reportsresearch/top-12-storage-companies-in-2015/
Share this Post