string(7) "English"

The End of the VMAX Era and the Rise of the All-Flash Array

Pure Storage got together with several hundred enthusiastic customers, partners and prospects last week in Vegas across the street from EMC World. The conversation was buzzing about how much keynote airtime next door was devoted to bashing Pure. Of course, most of this talk was about Pure’s direct competitor XtremIO and EMC’s latest AFA acquisition, DSSD (more on that below), as well as EMC’s shenanigans around offering a $1m challenge, for which you can read our response on here.

However, all this distracts from the real competition in the marketplace, which is between AFAs like Pure Storage’s FlashArray and VMAX!  Symmetrix/VMAX has been the gold standard for performance (a.k.a. Tier 1) storage for 20 years. VMAX competes with Hitachi, HP 3PAR, NetApp, IBM XIV, EMC’s own VNX, et al. for an approximately $15B annual spend ($60B in total addressable market). And VMAX remains one of EMC’s top two sellers (along with VMware).

Pure’s claim is that this VMAX-led market is tipping to AFAs over the course of the next refresh cycle (4-5 years). With the announcement of new hardware and software we’ll make later this week, the remaining technology puzzle pieces of native replication and multi-100TB scale will be in place to enable the enterprise flash revolution. (This week Pure Storage is also hosting the Flash Visionaries Summit, an enduser event showcasing flash learnings and best practices from leading endusers like LinkedIn, Workday, Shutterfly, and several more.) Pure is already seeing this transition in our installed base, with an increasing share of our customers swearing off any further purchases of VMAX or its competitors in favor of the AFA alternatives. Some of the most prescient have already gone 100% solid-state in their data centers. Others have accelerated the depreciation of hybrid storage to get to the all-flash future more quickly.

Why? The All-Flash Array recipe that Pure originated yields a product that is 10X faster, more space & power efficient, more reliable, and simpler, and yet typically costs less. Prior Tier 1 disk contenders for the VMAX crown simply did not offer the order of magnitude improvement necessary to unseat the market leader. As further proof points that AFAs are now capable of superseding mission-critical disk & hybrid storage, Pure Storage’s FlashArray

  • Delivers >99.999% uptime, as measured across our installed base (we shipped out 1,000th array last year), and does so with no required maintenance windows (upgrade, expansion, and hardware repairs can all be done non-disruptively and without performance impact);
  • Removes on average more than one million hours of latency per year between a business’ applications and the underlying storage (What could your business do with a million hours of greater information systems productivity next year?);
  • Saves enough in power, cooling and operating overhead in it’s first 4 years of deployment to pay back customers for the initial purchase price (If you’re still buying high-end disk arrays, be sure to negotiate a power rebate); and
  • Provides sufficiently high quality to warrant a NetPromoter Score™ of 76, one of the highest in the tech industry and 2-3X that of the traditional disk storage vendors.

With a $60B market at stake, you can expect the AFA competition to heat up dramatically this year and next. EMC launched XtremIO late last year, and yet also purchased Andy Bechtolshiem’s company DSSD. Chad Sakac offers an explanation for this investment here, but it seems to me that supremely low latency is more of a sweet spot for the in-memory crowd (SAP Hana) than it is the big data crowd (Hadoop) where cost is king. My own speculation is rather that EMC felt exposed with both VMAX and VNX moving into capacity storage and only one AFA in their portfolio: Is XtremIO the new VNX and DSSD the new VMAX? Or has the bet on XtremIO not panned out entirely as expected? It is also possible that EMC, an early investor in DSSD, acquired the company to keep it from a competitor, as was rumored to be the case with XtremIO. Either way, DSSD is a formidable team that was doing innovative work in both storage hardware and software. It will be interesting to watch how the product line develops within EMC. Whether DSSD remains a separate product or converges with XtremIO, EMC is virtually certain to remain Pure’s primary competitor in both the legacy VMAX/VNX category as well as the AFAs of the future.

Before your organization commits to four more years of performance (oxymoron) hybrid disk storage, I have a simple ask: Please try one of the AFAs out on your workload. The strong companies will let you do proofs of concept (POCs) on their dime, and Pure even offers a money back guarantee if you’re not delighted by the transformation all-flash storage enables in your business. The sooner your business joins the all-flash revolution and takes advantage of that disruption, the further ahead you will be relative to your competitors. Thanks to our getting this disruption right, a great team and great go to market partners, Pure Storage is growing gangbusters. And please, join us on Thursday as we pour more fuel on the rapid transition from Tier 1 disk to flash!