duquesne
Research Duquesne Advisory delivers in-depth analyses of Information and Communications Technologies, their implementations and their markets. Research is based on critical observation of the market by the analysts and their on-going contacts with the vendor community, together with hands-on, practical experience in consulting engagements.

EMC/VMware and Cisco data center alliance: competing for the future ?



EMC/VMware and Cisco data center alliance: competing for the future ?
The recently announced Virtual Computing Environment (VCE) coalition, associating EMC /VMware and Cisco, has been born at the beginning of a period of dramatic evolution in the data center infrastructure market. The VCE coalition claims to be based on “four cornerstones”: “technology innovations, unified customer engagement via integrated sales, services and support, Acadia (see below) and partner ecosystem leverage”. Its inception has already resulted in the offering of "Vblocks", which are standardised, pre-assembled sets of components from EMC, Cisco and VMware. At this point in time, Vblocks come in three different "sizes" and enable the very fast deployment of new-generation data centers. They will be marketed “independently” by each of the alliance members.

In addition, a Cisco/EMC joint-venture company, Acadia, is aimed at accelerating acceptance of these commonly engineered solutions by offering implementation and initial operations services. Acadia, in which Intel Corp. is a minority shareholder, proposes a "build, operate and transfer" (BOT) services model.

To understand this coalition, it needs to be put in the context of the four key influences which will drive change in IT, including the data center infrastructure market, in the 2010 – 2020 decade : 

1 - evolving customer demands, expectations and strategies,
2 - profound changes in application portfolios,
3 - continued technology innovation (and associated obsolescence of current  solutions),
4 - changes in the IT supply and consumption ecosystem.
 
In this dynamic context, the strategies of key infrastructure players must evolve; new approaches can be adopted, and competitive balance can be shifted over time. The EMC / Cisco / VMware alliance appears as one such new approach, with the potential to significantly redefine the data center market, but also with significant challenges to address.

Duquesne believes this alliance makes a lot of sense for Cisco, a leader in networks but a new entrant in server infrastructure. Partnering with the leaders in storage and virtualization reinforces the credibility of its offering, at a time when competition with server-centric vendors intensifies as network consolidation initiatives tend to shrink the data-center LAN footprint. For EMC, it reflects the need to add muscle to its own offering, as its main competitors in storage, IBM and HP, dominate the server component of the data center market and take the lead of many consolidation projects.

However, for both companies and also (even more in fact) for VMware, the challenge will be to succeed in a massively “coopetitive” market, where all competitors are also business partners…and where market shares will not shift very rapidly. Nevertheless, before VCE and Acadia, the cooperation areas and mutual interests between Cisco or EMC, on one hand, and IBM, HP and others, on the other hand, largely outweighed the competitive situations. This will change, and perceptions by competitors will change, driving more arms-length-like relationships. But for a long time ahead, both EMC and Cisco will still need to preserve the revenue stemming from cooperation with these competitors. To name just one example, IBM Global Services data centers are a large client for EMC storage products.

Although the alliance will not change the face of the worldwide data center market in the short-term, it enters the market at an appropriate time, and potentially can leverage the four above-mentioned “influences”.

1. Evolving customer demands, expectations and strategies

Large IT user organizations will continue to seek cost effectiveness through infrastructure optimization and rationalization. With virtualization now a broadly accepted technology, organizations will seek to further industrialize their infrastructure; at the same time, growing incentives to adopting “green” solutions will also encourage infrastructure modernization.

Data volumes and usage of information will continue to grow, driving a continued development of storage infrastructures and further requirements to effectively access those data. 

User organizations will definitely, yet cautiously, engage into “cloud computing” and the “private cloud” approach is likely to attract interest as it simplifies somewhat the daunting security issues posed by public cloud offerings. Indeed, cloud computing is less about introducing new technologies (even though these technologies are deployed in different, specific ways) than about provisioning and financing technology resources in a different way.

The combination of Cisco and EMC offerings in a packaged solution - reducing deployment time and offsetting integration issues from the Clients - makes the move potentially less painful, but at the expense of a significant degree of lock-in. Vblocks implement Cisco's Unified Computing System (UCS) "unified" architecture, which is proprietary (like other blade server infrastructure offerings, by the way).

While the alliance brings added credibility to both Cisco and EMC in the data center space, the main beneficiary is Cisco. VCE (and Acadia) will help promote acceptance of its proprietary UCS technology, while EMC "only" promotes its already established storage systems and management software at this time. Indeed, as exemplified by the more recent, but yet almost simultaneous announcement of another cooperation initiative linking (albeit more loosely) Cisco and VMware with NetApps to promote an offering labeled "Secure Shared Infrastructure" (SSI), Cisco has engaged in an aggressive effort to boost UCS acceptance in the marketplace. The VCE coalition, as well as SSI, are only pieces of the strategy: “Everything should work with UCS”.


2. Profound changes in application portfolios

As the extension of information systems from “internal nervous systems” to “external-oriented communication and collaboration channels” continues, the percentage of applications that can be hosted on standardized infrastructures (i.e. x86 server-based infrastructures) will continue to grow. Such applications will include reengineered or modernised "traditional" production applications, as well as packages or standard-development platform based "new" applications (Web 2.0, collaboration, communication, etc.).

Pre-assembled building blocks like Vblocks will be well-suited to this growing type of requirement. However, such building blocks will be better suited to new infrastructure deployments – especially those with “Greenfield” data centers”- while their cost-benefit ratio will be less favourable when integrating tightly with existing systems. In this case, traditional data center integration solutions will remain competitive.

Competition will be fierce in this building block market, however. While Cisco / EMC and HP will primarily compete on server and network infrastructure, IBM and Oracle/Sun will extend their offering further, integrating database and (later) application software. IBM will continue to lead the charge in extensive data center value-added services.

Duquesne believes that while a big minority of large IT user organisations will definitely seek one-stop-shopping solutions (exemplified by IBM's offerings and now - to a degree - by Oracle's or Acadia's), others will seek cost and service optimisation by selecting relevant components from a limited set of suppliers and maintaining a sufficient level of vendor (and technology) independence.  Whilst the current Vblock offerings includes VMware VSphere, RSA’s (the security division of EMC) security solutions, and Ionix (EMC’s storage management solution), we expect a number of clients to stick with their current security and management choices. To expand business in the Windows server market, we also anticipate that Cisco and EMC will have to support Microsoft Hyper-V as an alternative to VMware.

3. Technology innovation and evolution

Cost effectiveness, better environmental footprint and increased performance will drive a steady renewal of IT infrastructures. With energy continuing to grow as a percentage of data center operations budgets, replacing data center facilities as well as server and storage infrastructures with new, more energy-effective solutions will increasingly prove more cost-effective, hence more attractive than incremental improvements to existing sites and equipments. However, with a large percentage of user organizations seeking IT capital expenses reduction or control, this will result in faster growth for data center and “cloud” services providers.

Duquesne believes that market demand for storage connectivity in 2010 and beyond will evolve away from costly, storage network dedicated technologies such as Fiber Channel towards iSCSI on carrier-class Ethernet, with SAS and SATA being the disk interface standards. The SAN will increasingly use the general data center network infrastructure, either using dedicated physical links or relying on proper quality of service management solutions to share the bandwidth with other types of traffic. As a side remark, we believe that other data-center specific technologies such as Infiniband will remain confined to niche usages, as low latency Ethernet technologies narrow the latency gap,  while the cost difference expands (including a much higher energy consumption on the Infiniband side).

The data center network of the future will therefore rely on three main connectivity layers:
    - in-memory communication between virtual machines (implemented at the virtualisation engine (aka hypervisor) level), 
    - a very high performance "chassis layer" connecting blade servers and other "blades" such as switching equipments (this layer will actually "collapse" a large part of the previous data center LAN),
    - carrier-class (no loss, low latency) high speed (10GB and beyond) Ethernet to connect the building blocks (and other legacy equipments) together. Users will arbitrate between more or less "blade-intensive" infrastructures based on their need for density and performance on one side, and the desire to avoid vendor lock-in on the other. 

EMC and Cisco both have a proven track record in integrating new technologies in their offerings; therefore we do not expect technology access and integration will be an issue in their competition with leader server vendors in the standards-based part of the market. 

4. Changes in the IT supply and consumption ecosystem

Duquesne expects the structure of the data center investment market to evolve over the decade, with service providers representing a growing percentage of total infrastructure spending.  This is a truly major opportunity for Cisco, as most service providers (especially internet-based providers and telco operators) are already huge Cisco customers, where the legitimacy of the joint offering in the server space will be easier to establish than in traditional, corporate “private” data centers.

With these "traditional" clients, gaining mindshare versus the large server vendors/integrators (IBM, HP) will prove more costly and time-consuming. It will also require more services, as many of these clients will either not have the capability and maturity, or not have the desire, to integrate such building blocks by themselves.

The Cisco /EMC / VMware coalition, in its current state, remains limited in scope, when compared to the offerings of IBM or HP. We believe the true challenge will come on the services front, where the capabilities of the alliance (and of both firms, indeed) significantly lag those of IBM or HP. (EMC’s service business is only 20% of revenue, including maintenance!). And the current resources allocated to Acadia itself are even more limited. In addition, as discussed above, “cloud computing” is very much about funding and financing IT resources. Matching the financial services capabilities of IBM and HP will be another significant major challenge.

Even assuming the two firms intend to develop their service activities (which would in all cases take time and require profound cultural changes), they will need at the same time to create a significant ecosystem of integrators and service providers, beyond and above the capabilities of their current indirect channel partners. This also will take time and effort. 

Finally, as they evolve from “specialty” vendors (in networks and storage, respectively) into core data center infrastructure global providers, Cisco and EMC / VMware will need to establish true customer intimacy at the highest level – beyond their current excellent credibility among technical experts and professionals in their respective fields.

Bottom Line

Overall, Duquesne believes the VCE coalition (and Acadia) is only a first, limited (in its current configuration) step towards the common objective pursued by Cisco and EMC/VMWare: establishing themselves as key full scope players in the "new" data centers. It will enable the three partner companies to acquire experience in deploying the new solutions. However, further steps will be required for the objective to be attained - steps that may be taken separately or jointly by the three firms. Duquesne expects both EMC and Cisco to continue expanding their own alliances with other players in the renewed technology ecosystem.

Meanwhile, from a customer perspective, the VCE coalition is a welcome addition to the competitive data center technology landscape. It represents a particularly credible option when deploying greenfield, "private cloud"-oriented data centers. Users, however, should carefully weigh the benefits of simplicity and speed versus the significant degree of vendor lock-in they will have to accept.

Thursday, March 4th 2010
Duquesne Advisory
Newsletter To subscribe to the Duquesne Advisory Newsletter, please enter your e-mail address.

Duquesne Advisory

Duquesne Advisory Ltd is a European firm, headquartered in the UK, dedicated to researching, understanding and advising clients worldwide on opportunities and trends in Information and Communications technology.

Research

Duquesne Advisory delivers in-depth analyses of Information and Communications Technologies, their implementations and their markets. Research is based on critical observation of the market by the analysts and their on-going contacts with the vendor community, together with hands-on, practical experience in consulting engagements.

Consulting

The analysts of Duquesne Advisory leverage the Firm’s ongoing market and technology research to undertake high added value consulting engagements for both ICT users and ICT providers. Focused on client service, their approach is rigorous and methodical, and at the same time pragmatic and operational.