The last week has seen quite a few developments in the competitive landscape for the storage industry, clarifying the market positioning of nearly every major player in this space.  The horses are on the track, lined up for the ‘flash storage derby’, and all of the bets are in. Wikibon Founder and CTO David Floyer set the field and handicapped the odds for us.  The players?  EMC, Dell, Hitachi, HP, IBM, NetApp. The strategy?  There are are two clear strategies being used by the major players to dominate the flash storage market: building on an existing platform vs. starting fresh.

According to Floyer, starting from scratch means the “emphasis has been on designing a system to take the maximum advantage of flash, focusing on the specific areas where flash makes a difference in, and focusing on making it cost competitive.”

Now the race is split with three players in each camp. Here are the companies in the building on an existing platform camp: Dell, Hitachi, HP.  And for the companies starting “fresh” we have: EMC, IBM, NetApp.

Floyer details the two kinds of “horses” in the Flash Derby during a recent apperance on our Live NewsDesk Show with Kristin Feledy.  See the video below for his full analysis.  We highlight the primary characteristics of each type of “horse” in this Flash Derby after the jump.

It’s clearly a two ‘type’ horse race: building flash on existing infrastructure vs. building flash anew. The flash storage race is crystal clear now, so who would you bet on?

We’ve outlined the two types of horses in this race below, detailing the top market players and which type of horse they are.  The differing strategies of leveraging existing tech or starting from scratch will converge on the battlefield of tomorrow’s data center, where disruptive tech must prove its worth in innovation, and mature tech must maintain its stance by standing on experience and foresight.

Those building on an existing platform

.

These traditional vendors (Dell, Hitachi, HP), have developed a strategy to lay flash on top of their own technology. The question is simple: can these traditional vendors leverage their existing architecture to not only support flash, but offer a better solution?

Here is an example of how one of these vendors would defend the ‘building on an existing platform’ side: HP’s existing architecture is a mature product that took a decade to build out. A disadvantage to everyone else spinning up new stacks is the lack of maturity there. HP has a full storage stack, and while 3PAR’s storage service is expensive, it’s still one of the best.

Here’s some bullet points on each of the three major players building flash into existing storage infrastructure:

• Dell: Compellent Data Progression
  • Compellent’s data progression is top down, it comes in on flash and goes down through the stack. Dell has built a flash only device with this architecture. Key advantage: Connects to a very dense disc layer.
• Hitachi: HDS (Hitachi Data Systems)
  • HDS’s SSD’s use the existing stack (same as Compellent) and tier-storage. Key advantage: Hitachi is using their existing stack as is using amendments to improve — you can integrate it with the rest of the systems.
• HP: 3PAR
  • 4-way processor that is extremely efficient in power, cooling and spacing. Bullish on software-defined storage. Key advantage: Ability to connect everything through peer motion – allows you to move dynamically sets of volume from one array to another. Useful feature because whether up to it or down form it, don’t have to take hits in downtime.

While HP, Dell and Hitachi are betting on the technology they’ve built, acquired and integrated over the years, anticipating a disruptive change to the storage space, others have been agitating the pot, hastening the data center’s evolution.  These disruptors could be startups, or fresh initiatives within a traditional vendor offering.  Either way, the storage industry at large is rushing to make the changes necessary for immediate survival and long-term success.

Those starting fresh

.

EMC, IBM and NetApp are developing from scratch a system that takes advantage of flash focusing on the specific efforts where flash makes the biggest difference: compression, deduplication, metro-clustering. The systems they’re building offer better performance of flash and utilization of flash.

• EMC: XtremIO
  • EMC is clearly not defending old technology buy the XtremIO purchase.    With a pure flash-only array, with XtremIO technology in its Thunder and Isilon products, EMC can provide a platform for Greenplum analytics to run on.
• IBM: TMS (Texas Memory Systems)
  • Has been saying Flash is going to revolution storage architecture for a long time without doing a whole lot in the space. Now, flash-optimized capabilities is part of their smarter storage strategy. IBM went the route of acquiring the capabilities for market share through the TMS acquisition and starting fresh with a good thoroughbred in all-flash arrays.
• NetApp: EF540 Flash Array
  • Purpose-built FlashRay product family. Very rich set of features that is focused on flash and flash-only startups (PureStorage, StorageIO). Focuses on deduplication. EF540 is high-performance and focuses on “the fat middle”.