A week ago, rumors swirled about EMC buying XtremIO. There is still good reason to believe that the acquisition will happen. I spoke with analysts over the past week who helped me get this story together. This represents their perspectives and what I have learned in my own research.

XtremIO resembles Solid Fire, Pure Storage and even Violin Memory to a certain extent. All flash, built from the ground up using data storage reduction techniques such as inline deduplication to lower the high costs.

As has been discussed many times, the bottleneck in today’s systems design is mechanical disk. Storage is the key pain point for application developers and limits the acceleration and agility of IT. All-flash arrays are coming fast and this category of system will dominate in the long-term for active data. Spinning disk will be relegated to storing data that is infrequently accessed and doesn’t require high performance.

So in the context of these rumors – which companies have the most to lose from the disruption that is flash? The established giants, of course, namely: EMC, NetApp, HP, IBM, Dell and Oracle.

EMC’s approach, up until it announced VF Cache, was to place flash as a device into a disk array system. The bottleneck here was still the disk controller, which was designed to support spinning disk. EMC’s VF Cache is an attempt to extend the flash hierarchy and bring storage closer to the processor…and manage that hierarchy where flash is the highest performance tier.

NetApp provides another example. It has a different strategy using a flash-cache type of approach and has shunned tiering. Nonetheless, both companies have a vested interest in designing or acquiring companies that have developed all flash technology with the associated management capabilities.

The Israel Connection

XtremIO is an Israel-based company. Isreal-based companies, for the most part, are not great at selling – what that they are great at engineering. There are some exceptions but generally Israel is an incubator, not a threat to develop the next EMC or NetApp. As a result they often make better acquisition candidates because they don’t have a strong channel and don’t covet doing an IPO such as Fusion-io or Violin. Companies like EMC and NetApp, that have big sales channels can hence pick up small Isreali startups for less money. But they’re not buying sales, they’re buying technology.

The Analyst View

Reported estimates put a sticker price of $400 million for the company. That’s comparable to Solidfire and Pure Storage. Violin is too rich at $800M+ and is eyeing an IPO. It’s also plausible that NetApp is looking at the company and even though, like Data Domain, EMC will win if it wants the company, NetApp could be stalking XtremIO to learn more while XtremIO is in sell mode. NetApp could also be bidding the price up.

Bottom Line: The flash/all-flash array movement is here to stay and the big whales are behind. They have lots to lose. The storage hierarchy is shifting and the role of spinning disk is diminishing as evidence by the small number of suppliers that make spinning disk. There used to be 50-70 suppliers – today there are two to three. Spinning disk is simply not strategic anymore and flash and the management of the flash hierarchy is.

Here’s a really good article from Wikibon’s David Floyer on the overall topic. In it he includes a market forecast.

As  Floyer’s numbers point out, flash is and will be faster, cheaper and consume lower power. To the extent the reliability can be demonstrated all-flash arrays will kill spinning disk for all active data.