Daily Archives: June 15, 2010

Dell Reloads in Mid-Market Data Center

Last week, Dell announced a fusillade of products for small- and medium-sized enterprises looking to benefit from converged, virtualized data centers.

Depending on one’s vantage point, Dell proactively announced the products to offer its mid-sized enterprise customers interoperable solutions that will allow them to derive efficiencies from data-center automation;  or it made the announcement reactively, in a bid to preclude incursions into its installed base by Cisco, IBM, HP, and perhaps even Oracle, which has yet to play the data-center-hardware hand that it was dealt in its marathon acquisition of Sun Microsystems.

In receiving an update yesterday from Brian Payne, director of Dell PowerEdge servers, and Mike Roberts, senior manager of Dell PowerEdge servers, I was struck by how much emphasis the Dell spokesmen placed on two key themes: openness and innovation.

For Dell, architectural openness is defined by interoperability, adherence to industry standards, and customer freedom from proprietary lock-in. Dell draws a distinction between its interoperable approach to data-center networking and the proprietary offerings of Cisco and, increasingly, HP. Dell contends that customers that adopt  converged data-center solutions from HP or Cisco — encompassing servers, storage, networking, and virtualization — could find themselves tied to a vendor that stops innovating. For those customers, the result could be competitive disadvantage, especially if their counterparts patronize vendors  — for instance, Dell — that offer an interoperable, open model.

This leads to a discussion of innovation. Dell is at pains to stress that it has innovated and continues to innovate in the data center. Indeed, while there’s nothing revolutionary or dramatically disruptive in Dell’s new slate of product announcements, the company is making noteworthy advances in its server architectures, its storage offerings, its management software, and its support for virtualization. It’s also innovating, through its Dell Business Ready Configurations, in offering preconfigured solution bundles to mid-size enterprises in target vertical markets.

Although Dell suffers from a brand-image hangover that has proven difficult to shake,  the company has escaped from the ghetto of white-label box vendors. To be sure, Dell still has chapters to write in its data-center narrative, but it is proving adept at devising and deploying viable technical architectures and business solutions for its target markets.

In that respect, what Rob Enderle, principal analyst at the eponymous Enderle Group, told Channel Insider rings true:

“The market likes choice and specialization. No one vendor, since IBM owned this market, has been able to be expert enough at all business sizes and types providing room for each vendor to specialize and carve out a market.”

“Dell tends to favor firms who want to do much of the work themselves, aren’t particularly interested in global services, and want a hardware vendor who is at arm’s length from software to avoid lock-in. There appear to be enough of those folks to sustain Dell.”

I generally agree. Moreover, I think a case can be made that those customers, once they’ve made a significant data-center buying decision, are unlikely to switch vendors unless they’re given a compelling reason to do so. Usually, though not necessarily, that impetus would involve their incumbent vendor falling woefully behind the innovation curve over a sustained period.

Dell is cognizant of the risk, which explains why the company is pushing the innovation theme so forcefully. It wants customers to understand that its interoperable converged data center doesn’t involve an innovation tradeoff in relation to alternatives from IBM, HP, and Cisco.

Accordingly, Dell draws attention to the fact that its new blade-server hardware features the latest industry-standard microprocessors from Intel (in the PowerEdge M710HD) and AMD (in the PowerEdge R715), not to mention an interesting utilization of general-purpose GPUs in its PowerEdge M610x.

Similarly, Dell cites automated data tiering and performance improvements in its EqualLogic PS6000XVS and PS6010XVS storage arrays. It also talks up the performance advances in its PowerVault MD3200 and PowerVault MD3200i storage arrays.

On the networking side, with the release of PowerConnect-J series of products, the first offerings derived from Dell’s OEM agreement with Juniper, Dell emphasizes that its customers can rely on Dell’s networking partnerships to ensure that they don’t suffer from Cisco envy. There is a similar message in Dell’s extension of the PowerConnect B-Series of chassis-based switches OEMed from Brocade. which recently gave its own notice that it has its head and heart back in the enterprise-networking fight.

Dell also draws attention to energy-efficiency enhancements delivered in its M1000e blade-server chassis, and it notes systems-management updates to its Lifecycle Controller, Chassis Management Controller, and Integrated Dell Remote Access Controller (iDRAC). Yes, a lot of this is rustic meat and potatoes, but it’s all part of the data-center buffet, and Dell needs to demonstrate that it hasn’t forgotten to provide  a full menu.

When I spoke with Dell, I got the feeling that it fears Cisco most of all. IBM plays upmarket, mostly out of Dell’s neighborhood, and HP is a known commodity — in more ways than one — perhaps with a reputation for enterprise innovation that is no longer warranted under the grim cost-cutting scythe of Mark Hurd’s technocrats of doom.

Cisco, though, seems to command Dell’s full attention. There appears to be a belief within Dell that Cisco won’t be content to spread its Unified Computing System (UCS) for data centers exclusively to high-end enterprises and cloud-based service providers. That assumption is probably correct. Dell has reason to be concerned.

Then again, healthy paranoia never hurt anybody. If concern about Cisco keeps Dell focused on delivering solutions to its core customers in the middle market, the preoccupation will have been a positive stimulus.

As a company, though, Dell might have a better chance defending its turf if it put more resources into its SME and enterprise strategies and product portfolios and proportionally fewer resources into consumer markets, where it seems destined to lose market share and squander brand equity.

Learning from F5’s Success Against Cisco

Earlier today while perusing the morning chatter on Twitter, I learned via Brad Reese that market-research firm Infonetics says F5 Networks has opened a commanding nearly 20-point lead over Cisco Systems in the market for application-delivery controllers (ADCs).

Citing the Infonetics data, which reaches as far as the first quarter of this year, Reese professes astonishment that F5 has surged so far ahead of Cisco after being in a neck-and-neck battle with the networking giant as recently as the fourth quarter of 2008.

I can’t say that I’m surprised at F5’s success. Despite Cisco making several acquisitions over the years in the load-balancing and application-delivery markets, it never has been able to find the silver bullet to smite F5.

We can see now that load balancing and its descendants were markets that got away from Cisco. If you look back into the ancient history of networking — yes, let’s go back as far as 1996, shall we?  — we see that Cisco first tossed its  LocalDirector into the ring to dispose of F5.

After LocalDirector was discontinued in 2003, Cisco turned to technology it had obtained as a result of its extravagant acquisition of ArrowPoint Communications, for a whopping $5.7 billion, in 2000. Before the discontinued of LocalDirector, Cisco went through an uncomfortable period where LocalDirector, technology from ArrowPoint, and other in-house technologies jostled for dominance in Cisco’s load-balancing portfolio. It wasn’t pretty, and it was confused for field-sales representatives and channel partners alike.

At this point in my narrative, I will digress, but stay with me.

In retrospect, Cisco might claim, given the obscene valuations of public and private companies back then, that its ArrowPoint purchase wasn’t as crazy as it looks from our current perspective. To support its point, Cisco might cite Nortel’s acquisition of Alteon WebSystems, an ArrowPoint competitor, for stock initially valued at $7.8 billion (but which was worth less by the time the deal closed).

But citing Nortel’s acquisitions as a rationalization for one’s own corporate debauchery probably isn’t the strongest defense. As a loquacious lawyer would say, that sort of rationalization might explain one’s actions, but it does not excuse them. Besides, Nortel eventually sold its Alteon assets to Radware for $17.65 million, which tells you all you need to know about why Nortel is a defunct company.

Returning from my historical perambulations, I want to draw a couple of inferences from the trip down memory lane. First, all those companies — not just Cisco, but Radware, ArrowPoint, Alteon (and later Nortel) — were F5 competitors. Second, once upon a time, given the erstwhile valuations of the likes of ArrowPoint and Alteon, the market attached whopping value to the space and felt it was ripe for big-vendor consolidation and a changing of the market-leadership guard.

But it didn’t happen. Despite all the big vendors buying companies around it — Intel also acquired a company called iPivot, for $500 million in 1999, that was trying to solve server-bottleneck problems — F5 more than held its own. And it continued to do so, even though Cisco has kept taking runs at it.

So, considering the big picture, why has F5 succeeded against Cisco where so many others have failed?

One of the reasons, and I have seen and heard others mention it, is focus. Unlike some vendors who’ve failed miserably against Cisco over the years — Nortel, the pre-Chinese incarnation of 3Com, Cabletron and its scattered progeny — F5 had market discipline, focus, and unswerving resolve. It was not blinded by hubris or delusions of grandeur. The company went deep rather than wide, stuck to what it knew best, and worked hard at staying close to its customers and building the best products on the market.

When F5 has chosen to expand into new markets, it has not done so hastily. The company has taken a measured approach, adding products purposefully and making acquisitions (most of the time) of the tuck-in variety, ensuring that such purchases are accretive in the near term. This was all part and parcel of F5’s business and market discipline.

What else? I think the company’s channel-sales structure was carefully planned and well built, drawing from best practices in networking and other sectors. After learning some hard lessons, F5 stressed quality over quantity (as in number of) partners. F5 constructed a sales-support apparatus that could scale effectively, providing all the necessary backing that channel partners needed to prosper.

In technology partnerships, F5 has been similarly focused. Look, for example, at its application-related partnerships with Microsoft, SAP, and Oracle. F5 knows where BIG-IP plays in those application environments, and it has derived considerable value from providing proven and innovative solutions that help customers run those applications more productively.  Again, discipline has been key. The same holds true for the company’s other technology partnerships across areas such as infrastructure, management, security, telecommunications, and virtualization.

Its solution sets and vertical-market programs have grown in a similarly deliberate manner.

Another important consideration — and, again, it’s something others have mentioned — is its community outreach to customers, IT professionals, and developers. In that regard, F5’s DevCentral has been indispensable. A lot of attention and resources are invested in DevCentral — and it shows. The site went through a refresh earlier this year, but it invariably has provided a welcoming forum for collaboration and discussion not only relating to F5’s products but also to some of the broader data-center challenges faced by IT professionals.

F5 demonstrates that it’s possible to compete and win against Cisco. As Cisco extends itself into market adjacencies, it is advancing into areas that are new to it but that often already have incumbent vendors. Those companies — in the smart grid, in next-generation IP video services, in digital signage, and in many other areas besides — should study the F5 playbook. It offers practical guidance on how to keep the giant at bay.