HDS's Hard-Wired Vision

Over the past few years, Hitachi Data Systems (HDS) clawed its way to being one of the leaders in the high-end RAID array business by building enormous, powerful systems that chewed up and spit out competing boxes.

But the world has changed around HDS. Software and services are becoming critical pieces of the puzzle in winning business. And because HDS's focus is still on building bigger-and-faster storage hardware, it's in serious danger of slipping behind its more well-rounded competitors, such as EMC Corp. (NYSE: EMC) and IBM Corp. (NYSE: IBM).

Let's take software first. What, exactly, is HDS's storage software strategy? Well, it provides array-based software like its competitors. But beyond that, it's not at all clear where HDS thinks it should go.

In an interview about a year ago, HDS president and COO Dave Roberson boasted that Hitachi had committed $500 million to software engineering over the next three years. But he also told us that HDS doesn't "want to reinvent technology that other companies have already brought to market. There's no point for me to do that, because it's just a waste of my R&D dollars. So we're happy to collaborate with Veritas, with IBM with whoever is out there we need to collaborate with – because we're going to build these industry-standard solutions." (See our interview with Roberson here.)

Since then, the lack of much action on this front by HDS shows that Roberson has passed on trying to develop the company into a more serious storage software player.HDS has "collaborated" with CommVault Systems Inc. and InterSAN Inc., with reseller agreements announced last fall (see HDS Resells InterSAN, CommVault). It announced a "collaboration" with Oracle Corp. (Nasdaq: ORCL) earlier this year (see Hitachi Gets Tight With Oracle). All the while, HDS has delivered just one new piece of software it developed itself, HiCommand Tuning Manager, a proprietary utility that is supposed to optimize databases running on HDS arrays. Then this April, it finally hired someone – Jack Domme, who was decamping from the quickly dying StorageNetworks – to head up its software division (see HDS Hires VP of Software).

All told, it's not a very impressive list of milestones. Another telling detail is that HDS has not articulated any clear strategy relating to storage virtualization services, whereas EMC, IBM, and Hewlett-Packard Co. (NYSE: HPQ) each have. My overall sense of HDS's software strategy is this: When you're makin' bacon selling spindles, why worry about tomorrow?

By contrast, EMC has seen the future – and a very big part of it is software. Two years ago, Hopkinton reorganized to address both storage systems and open software. This year, it acquired data backup and archive specialist Legato Systems Inc. (Nasdaq: LGTO). EMC has set explicit goals, too: It plans to generate 30 percent of its total revenues from software by the end of 2004 (see EMC Salivates Over Software).

Granted, EMC hasn't seen dramatic success in the open software market yet. And there are risks with any deal the size of the Legato acquisition (see EMC's Soft Spot?). But EMC has positioned itself well here, whereas HDS has simply paid it lip service.

Hardware HeadsHere's a good case in point that illustrates HDS's hardware-über-alles mentality: This summer EMC announced enhancements to its Symmetrix DMX family, boosting its maximum capacity to 576 drives. It rolled out new software features, including a new version of Symmetrix Remote Data Facility (SRDF) that provides asynchronous replication and a new version of the Symmetrix operating system that allows for nondisruptive upgrades. EMC also said it would offer iSCSI and Ficon connectivity for the DMX. In discussing the new high-end DMX3000, CEO Joe Tucci said that since the initial launch of the DMX in February "absolutely zero accounts" have requested a system with more than 400 drives (see EMC Debuts DMX, Part Deux).

Guess what HDS's response was to this? Hey, we have customers that need more than 400 drives in a single system! HDS's marketing team either wasn't able to think outside its hardware cage or its product offerings didn't allow them to.

And HDS continues to neglect software at the expense of hardware features. Next month the company is expected to roll out a new member of its midrange Thunder series, the 9580V (see HDS to Crank Up Thunder). The system reportedly will double the cache available – to 8 Gbytes – and double the front-end connectivity of the system, compared with the 9570V, to a maximum of eight Fibre Channel ports. But is there any new software? Evidently nothing to write home about, aside from an app called HiCopy that will be able to replicate data across HDS's own arrays.

It's not just the absence of a credible software play that's holding HDS back. Professional services are also an increasingly important factor for success in this market. And HDS has not been able to catch up to its competitors on this front, either.

HDS only hired a general manager to oversee professional services less than a year ago (see HDS Appoints Services GM). The size of its services organization pales in comparison to those of EMC, IBM, HP, or Storage Technology Corp. (StorageTek) (NYSE: STK). According to IDC, HDS generated $145 million in storage services revenues in 2002. IBM pulled in more than 30 times that, with $4.7 billion. EMC tallied $1.1 billion; HP and Compaq had $1.9 billion; and StorageTek had $763 million (see HP Services Stay Self Centered).Perhaps not surprisingly, when Byte and Switch Insider researchers contacted HDS's professional services group about providing information for our October report comparing the offerings of more than a dozen industry players, the company declined to participate. Given the relative immaturity of the HDS services group, it would not have compared favorably to larger and stronger competitors.

The bottom line is that the prime directive at HDS isn't to become a more strategic supplier to its customers – it's to increase the attach rate of its Lightning storage arrays. The company is still counting disk drives as the measure of success.

Moreover, in recent months, the company's main strategy for boosting its presence in the SAN market apparently has centered on a slew of new spiffs designed to motivate its sales force. This approach is not designed to solve customer problems. It's a short-term grab for market share.

Memo to HDS: You're fighting yesterday's battles. Customers aren't interested only in the biggest/fastest/meanest hardware that money can buy anymore. They want complete solutions – which encompass storage hardware, software, and services – and unless HDS can deliver them, the company is going to fall even further behind the rest of the pack.

— Todd Spangler, US Editor, Byte and SwitchThe current Byte and Switch Insider report – "Midrange Storage Systems" – is available here