On today's (10/25/2007) EMC earnings conference call, an investor asked a question about the impact of storage virtualization on array purchasing. The questioner noted that IT buyers are looking at storage virtualization as a way to consolidate array capacities and the question essentially asked how this might affect EMC's array business.
Joe Tucci, EMC's Chairman and CEO responded by turning a somewhat negative implication into a positive by identifying three virtualization opportunities for customers to drive value, in order of importance, including:
- File virtualization using global namespace technology;
- Thin provisioning (EMC calls this Virtual Provisioning) as applied to existing infrastructure;
- SAN virtualization to address data movement between tiers.
Tucci's point on #1 is that file growth (unstructured data) far outpaces the growth of structured data and applying virtualization to this problem will drive near term customer value by improving the efficiency of under-utilized storage and easing management complexity.
On #2, Tucci pointed out that EMC is shipping thin provisioning on its Celerra NS Series (file based) and plans to have 'Virtual Provisioning' (its version of thin provisioning) on its entire product line (including block-based) early next year, which again will help boost utilization.
Tucci suggested that on #3, SAN virtualization, that customers have done a good job driving up utilization with existing SAN technologies and virtualizing SAN islands with EMC's Invista product will focus on easing migration pain but won't do much for utilization.
This seemed like a pretty reasonable way to frame the issue although observers will note that EMC is behind on delivering portions of key technologies mentioned, including block-based thin provisioning. Additionally, the company continues to underwhelm with Invista progress (although expectations are high for a renewed push in the near term).
In a later question about the effectiveness of startups (which remained un-named on the call) in delivering some of these technologies, Tucci indicated that EMC has strong responses for thin provisioning (3PAR), iSCSI (EqualLogic), spindown (Copan) and others. Tucci acknowledged that startups have done a good job focusing and beating EMC to market with these technologies but suggested EMC's strategy of integrating point technologies into EMC's robust product line will eventually overwhelm the value proposition of point products from startups. He also put forth the promise that EMC's products will ultimately be much more robust than startup technologies.
Tucci's assessment of the value to customers is probably about right although reasonable people could argue about the order of importance. To the extent that EMC executes and delivers on its promise of integrating robust and high quality function in its product lines, customers should strongly consider sticking with EMC and avoid switching costs. Historically, this approach has worked for the likes of IBM, CA, Oracle, Cisco and Microsoft, although with these firms and EMC alike, integration across product lines has always been wanting.
For customers the question remains how and when will EMC deliver on these innovations. EMC has been open about its intentions, which freezes the market on the competition but also puts the pressure on EMC to deliver.
Action Item: Action Item: Customers should begin evaluating best-of-breed function from upstarts in their labs to familiarize themselves with the benefits of innovative approaches from emerging companies. This will enable users to more accurately evaluate EMC's functionality and negotiate based on value comparisons with leading industry innovators.
Footnotes: Editor's Note: As of December 2008, EMC continues to struggle with Invista adoption and seems to be de-emphasizing the product. While the company has announced and is evidently shipping Virtual Provisioning it is seems this capability is more of a competitive 'checkbox' item versus a feature customers fundamentally require from EMC.