This document is an updated version of the document published on 26 March 2008.

Every vendor in the Gartner 2008 Campus LAN (Global) Magic Quadrant provides viable, well-supported LAN infrastructures for mainstream requirements. It is imperative that enterprises evaluate competitive offerings to ensure that they balance required features with the total cost of ownership for the anticipated five-to-seven-year life span of new switches. Gartner continues to recommend that most enterprises provision 10/100 switches with Power over Ethernet (PoE) at the edge of the network, with appropriately sized two- or three-tier architectures for overall scalability. Major LAN purchases and upgrades should not take place without a full competitive review.

The LAN switching market is mature; however, its importance is increasing, as enterprises move toward a secure, collaborative workplace and Ethernet takes over more data center functions. During the past two years, four key, largely product-related attributes have emerged that drive nearly every LAN upgrade:

• Support for an open, standards-based platform for convergence applications
• Ability to embed security in the LAN infrastructure
• Support for increasingly large, scalable data centers
• Ability to drive innovation to support a long-lived, cost-effective infrastructure

Although the market is mature, we are seeing significant market shifts. Cisco is still the dominant vendor, especially when looked at from a revenue perspective; however, its port market share is less than 55%, which means that 45% of enterprise Ethernet ports are not from Cisco. Looking at the alternative vendors, we are starting to see a stratification, with HP ProCurve breaking away from the pack. Nortel and Foundry Networks are also separating themselves from the remaining vendors. We are seeing varying approaches in enterprise designs. In one case, large enterprises are deploying Cisco in the core and HP ProCurve (or possibly H3C in Asia) as an edge switch. Foundry is winning high-end data center business, as well as end-to-end LAN switching solutions, while Nortel is capitalizing on complete voice and data converged solutions. Enterasys and Extreme Networks are showing signs of recovery, and have attractive offerings for secure, converged networks. Alcatel-Lucent has also executed well in Europe.

3Com is still in a state of confusion, despite strong aggregate revenue from its own business, H3C sales in developing countries and OEM sales to Huawei. Assuming a completion of the proposed sale to Bain Capital, the re-emergence of Huawei as an investor would be a positive. However, we have concerns about the interim uncertainty of the deal and likely changes in the future management team and new direction.

Because nearly every vendor can provide required infrastructure capabilities, it is extremely important for network managers to evaluate the alternatives in the market. Beyond specific product criteria, cost and support capabilities should be primary aspects of LAN switching evaluations.

The LAN switching market is a long-standing, mature market, with roots reaching back to the early days of shared Ethernet 10BaseT networks of the late 1980s. Global revenue (for business-class products) in 2006 was more than $14 billion, increasing to $16 billion in 2007, making LAN infrastructure one of the largest network equipment expenditures in the enterprise. The market has evolved to one where products are expected to have a useful deployment life of at least five to seven years (see "Toolkit Tactical Guideline: New Enterprise Network Useful Life Recommendations"). Although mature, this market is not commodity-based. Significant innovation exists in the market at a technology and economic level; however, as detailed in the fourth criteria area, a significant percentage of the market should be looking for more cost-effective, easier-to-manage solutions to their infrastructure.

Many new and emerging decision criteria must be considered when looking at the market. Support for Internet Protocol (IP) telephony and related applications is a key area. Specific aspects of the required technology include a flexible PoE capability, rapid failover mechanisms, autoconfiguration, quality of service and negotiation of device requirements. Standards such as Link Layer Discover Protocol-Media Endpoint Discovery (LLDP-MED) and Institute of Electrical and Electronics Engineers 802.1ab have emerged as key capabilities to provide discovery and autoconfiguration of the LAN infrastructure to support IP phones and other similar devices. Extensions to support more-granular PoE configurations are also key to ensuring support in case of emergency. The standards have not caught up with rapid failover capabilities, so the vendors have innovated around and beyond Rapid Spanning Tree Protocol to ensure voice quality and sessions during switch or link failures.

On the security front, network access control is another area of interest in the market. We expect network access control to be a standard feature of switches in the future; however, a lack of standards means that integrated approaches are largely proprietary. Although network access control will become an important capability, embedded security will expand to provide more-complete protection, and will include technologies such as post admission control, threat containment and content security. New vendors are emerging with embedded network security as their core competency, and they are adding comprehensive switching features to compete at the edge of the network. The best example of one of these new vendors is ConSentry Networks. ConSentry is starting to get traction with a security switch with network access control, post-admission, policy-based user control, threat control and full L2-7 inspection and visibility in a 48 port switch with 10/100/1000 and PoE.

In the core of the network, we're also observing an evolution of requirements. Large data centers are becoming increasingly dependent on the network layer. Requirements for emerging data centers can involve thousands of Gigabit Ethernet server connections, with a requirement for a 10-gigabit interconnect between the switches. Another approach is to deploy a dedicated server rack switch with a 10-gigabit uplink into the core. To meet either of these high-end architectures, an individual core switch must support many hundreds of gigabit connections and have dense, wire-speed 10-gigabit capabilities. Switches need to have enough combined port density to allow switch interconnection without seriously reducing the number of devices connected to the switch. We are finding the high-end market stratified, with vendors such as Force10 Networks and Foundry emerging as the leading players in high-capacity, scalable architectures. The future availability of the Cisco Nexus 7xxx line will enable Cisco to compete more ably in the high-end data center market.

These new data center requirements are causing a new category of switches to emerge. It is no longer valid to think of one, core switching market, where there is a near total overlap between the connectivity required to support user connections and that required for connecting servers and storage in the data center. This split between enterprise connectivity and data center connectivity will become more prevalent. Already, we are seeing new vendors that — although they don't yet qualify for inclusion in the Global Campus LAN Magic Quadrant — are key innovators. Blade Network Technologies has emerged as the leading vendor for blade server connectivity, with a vision to expand to a broader data center role. Teak Systems is targeting the 10-gigabit blade server market, and will expand to other data center opportunities. In addition, Woven Systems is attempting to deliver a new data center fabric switch that significantly lowers latency, power and cost (see "Cool Vendors in Enterprise Communications, 2007").

The final key evaluation criterion is focused on the cyclical nature of the LAN market. A significant portion of the market hasn't used the sophisticated features available in many vendors' products, and now faces upgrades as their installed bases age. Organizations that are not driven by network innovation should focus on the life cycle costs of their infrastructure, including capital costs, warranty terms and conditions, and available services. Organizations that are upgrading infrastructure that is more than seven years old (that is, was acquired during the Y2K upgrade cycle) should focus on procuring a standard, rich feature set, but target cost reductions for capital and maintenance costs of at least 50%. The market has options for lower-cost switches (from a range of providers), comprehensive lifetime warranties and free software upgrades to allow for a more cost-effective economic model, without giving up the required level of functionality.

For the broader market, cost should be an elemental consideration when looking at LAN infrastructure. Many enterprises overprovision their LAN, especially at the edge. Enterprises should procure 10/100 edge switches with PoE for the majority of their user populations. Only functions with exceptional network requirements (such as video production engineering, medical imaging, geographic information systems, high-end computer-aided design and scientific research) would justify the use of gigabit to the desktop. The adjustment to purchase 10/100 as the standard connection would save more than 15% of overall LAN procurement and maintenance costs for the enterprise. Further fine-tuning of requirements, combined with competitive bidding, would reduce these costs by an additional 15% to 50%.

Vendor viability comes up in many discussions, given the large discrepancy between Cisco's LAN revenue and the rest of the market. However, there is no significant risk with any of the top seven vendors in this market.

Vendors in the Gartner 2008 Campus LAN (Global) Magic Quadrant need to demonstrate an ability to deliver enterprise-class solutions for LAN infrastructure to the global market. Market share should demonstrate at least 1% revenue or port share of the broad, enterprise-focused Ethernet market, or at least 5% in a significant market segment. Vendors included have the ability to provide a full campus LAN infrastructure, or have innovated in select areas to provide key and differentiated functionality.

No vendors were added. Huawei/H3C continues to be a large player in the market — predominantly in the Asia/Pacific region and other emerging markets of the developing world. With 3Com taking over full ownership of the Huawei-3Com joint venture, the pending sale of 3Com and Huawei's uncertain/fluctuating commitment to the enterprise market, there are too many uncertain elements to do a proper rating. In the future, if there is a clearer direction by Huawei, and delineation between the direction of 3Com and Huawei, then we will add the analysis.

Juniper announced its entry into the Ethernet switch market in January 2008 (see "Juniper Entry Needs More to Compete in Ethernet Switch Market"). Juniper will be added to the analysis when it meets our inclusion criteria.

No vendors were dropped.

The following provides some insight into the criteria Gartner uses when evaluating a vendor's ability to execute. At a high level, our analysis of ability to execute attempts to capture how well a vendor is performing across primary functional units of the business — product, sales/channels, marketing, service/support and financial.

Product evaluates vendors by looking at their overall portfolio, with a particular focus on the four attributes identified: convergence, embedded security, data center scalability and life cycle attributes. More emphasis was placed on capabilities that would apply in an open, multivendor application scenario, because many of these areas cross boundaries of the IT architecture, making proprietary protocols a problem.

Overall Viability looks at a vendor's investments in a specific market, its financial investments and capabilities, and its long-term viability.

Sales Execution/Pricing was weighted higher than others on the execution axis, because it combines an evaluation of the presales and go-to-market activities, as well as an analysis of the resulting pricing and solution to the enterprise. On presales activity, the evaluation focuses on the vendor and its channel's ability to deliver comprehensive LAN infrastructure solutions — especially those focused on the three technical areas of innovation: convergence, security and data center. The second aspect of this criterion includes our evaluation of the cost-effectiveness of the solutions for capital purchase and long-term maintenance.

Market Execution focuses on how the vendor is perceived in the market and how well its marketing programs are recognized. For LAN infrastructure, the evaluation focused on how well the vendor was able to influence the market around key messages and attributes related to the four key areas in the market today. An additional indicator for this criterion is how often Gartner clients consider a vendor as a possible supplier in a shortlist evaluation. The change in momentum in this indicator is particularly important.

Customer Experience looks at all aspects of the customer interaction, with a heavier weighting on post-sales service and support activities (see Table 1).

Evaluations for completeness of vision attempt to determine how well the vendor understands and is preparing for future market conditions, as well as for shaping the future market.

Market Understanding looks at the vendor's ability to look into the future and drive new ideas into product road maps and offerings. In this market, leadership in driving the four key areas into the product offering is a good example, and demonstrates key abilities in this area.

Marketing Strategy evaluates the ability of the vendor to influence the market through its messaging and marketing campaigns. Vendors that have incorporated the key LAN criteria discussion points in the industry have demonstrated an ability to use their marketing strategies to their advantage. Examples are Cisco's "self-defending network" marketing campaign and Force10's focus on scaling high-performance computing labs.

Offering Strategy evaluates how the vendor invests in R&D to continue to innovate in key areas and ensure that future products continue to evolve.

Business Model provides a view into the vendor's overall commitment to the market, and its willingness to continue to make the right investments across all aspects of the LAN business unit.

Innovation measures the vendor's ability to drive into new, related areas of LAN switching, and to help move not only its own business, but also the market.

Sales Strategy, Vertical/Industry Strategy and Geographic Strategy were not ranked, given the maturity and horizontal nature of the market (see Table 2).

A leader has demonstrated a sustained ability to meet the changing needs for mainstream LAN switching architectures. A leader also has an ability to shape the market and maintain strong relationships with its channels and customers, and has no major gaps in the portfolio.

A challenger has demonstrated sustained execution in the marketplace, and has clear, long-term viability in the market, but has not shown the ability to shape and transform the market.

A visionary has demonstrated an ability to increase features in its offering, to provide a unique and differentiated approach to the market. A visionary has innovated in one or more of the key areas of campus LAN technologies (such as convergence, security, data center and operational efficiency).

A niche player has a complete or near-complete product offering, but does not have strong go-to-market capabilities or innovation in its product offering. A niche player still has a viable product offering and, in some cases, will be an appropriate choice for large infrastructure deals.

• Assuming that the privatization deal with Bain Capital and Huawei can be finalized, a renewed commitment by Huawei to the 3Com product line would provide a level of financial stability and significant growth prospects in developing markets.
• This vendor has a cost-effective, broad product offering, from small workgroup switching to very large high-capacity core switches. 3Com's technology has the ability to stretch from the small- and midsize-business market to the largest enterprise.
• 3Com's move to an in-house services portfolio is a strong move to meet enterprise requirements.
• Through its ownership in H3C, this vendor has tremendous access to the Asian marketplace, for sales and development.
• An open-services networking strategy has the ability to provide innovative services to 3Com switch platforms in the future.

• Uncertainty over 3Com's future and the delays in getting U.S. Treasury Department Committee on Foreign Investment in the United States approval for the privatization deal with Bain Capital and Huawei have a negative effect on this vendor.
• There is expected uncertainty and another management team overhaul with the pending sale to Bain Capital and Huawei.
• 3Com's poor execution of the Huawei joint venture over the past two years resulted in varying support from Huawei in developing markets and a loss of momentum.
• Poor brand image in North America and Western Europe has prevented the vendor from taking advantage of renewed product development capabilities.

• Alcatel-Lucent has a strong portfolio of products with an operational focus. Its core architectures are robust, and it has good embedded security capabilities.
• OmniSwitch 6850L enables this vendor's customers to buy 10/100 today, but upgrade to gigabit in the future. This is an innovative feature that completely removes the debate concerning the deployment of Gigabit Ethernet.
• This vendor's renewed enterprise focus has led to an increased ability to execute. Strength in Western Europe drove strong revenue growth through 2007.
• Alcatel-Lucent generally offers a lower-cost alternative to some of the larger players.

• This is a small vendor in the enterprise data market, with a lack of resources in direct-touch roles.
• It still has little field presence in North America.
• There is little or no evidence of any benefit from the Alcatel/Lucent merger for the enterprise business.

• Cisco has a broad portfolio, with a variety of workgroup switch, aggregation and core switch platforms. The Catalyst 4500 was recently upgraded, bringing improved performance levels (but still not market-leading performance). The recently announced Nexus 7010 fills a major gap in Cisco's high end, and the Catalyst 65xx VSS is a nice operational improvement.
• This vendor has the most globally distributed support capabilities. For organizations that need local support in remote corners of the globe, especially outside major business centers, Cisco clearly has the most universal coverage.
• Increasingly strong capabilities in unified communications result in this vendor making investments to integrate end to end, rather than tackling problems as a LAN switch vendor.
• The ubiquitous availability of trained staff makes Cisco an easy choice for many organizations.

• Despite having a large portfolio, holes remain in Cisco's portfolio and approach:
  • There is a lack of a cost-effective stacking product.
  • The vendor is slow to add standard features for convergence, such as LLDP-MED (just added in September 2007).
• Cisco has support challenges when dealing with multivendor solutions, especially in voice and collaboration.
• A business model built on strong account control and a lack of competitive bidding has allowed this vendor to keep list prices high.

• Enterasys has the most tightly integrated security capabilities of any LAN switch vendor, including strong support for all IP telephony offerings.
• This vendor has successfully made the transition from a public to a privately owned company, has solidified its installed base and is starting to show the ability to win new customer accounts.
• It has migrated from often premium-priced offerings to a more competitive, but still feature-rich solution.
• Enterasys provides very strong service capabilities. Long-term and new customers cite customer support as a differentiator.

• It is one of the smaller players in the market.
• Despite the improved prospects shown in the field, Enterasys is still losing market share.
• This vendor's local coverage outside of major centers is lacking.

• Extreme Networks continues to execute well on its XOS strategy, and is introducing a series of features — internally developed and from third parties — taking advantage of its cross-platform and open control plane.
• In 2007, the X250 was introduced — a cost-effective, full-featured 10/100 stackable switch family — and the X150 stand-alone family was introduced for small locations. These introductions give Extreme a strong, end-to-end XOS-based portfolio.
• With its new, business-focused CEO, we believe that Extreme is at the start of executing a turnaround in its business fortunes.

• Although improving, this vendor must demonstrate consistent execution in marketing, sales and service.
• Although its revenue is stable, its market share is still declining.
• Extreme is one of the smaller players in the market, and needs to focus on aggressively building market share with its newly crafted portfolio and messaging.

• Force10 Networks' strong data center switching platforms provided a competitive level of density and performance for large-scale enterprise and Web 2.0 data centers.
• New platforms have resulted in this vendor's more complete portfolio of high-performance LAN switches from edge to core. The transition (and migration to the left on the Magic Quadrant) is a typical one for a smaller vendor, as it moves from having a single area of innovation to being a broader player that still needs to demonstrate a more complete execution and vision.
• Force10 has a strong technical road map, including significant investment in its FTOS software.

• This vendor has not been able to substantially break out of its niche market, and new, emerging vendors are poised to leapfrog its once-innovative offerings.
• Revenue declined during in 1H07, and Force10's financial future was an impediment to enterprise sales. However, 2H07 showed significant improvement, and the first signs that it is making progress with its transition.
• Force10's geographic presence is largely limited to major business centers.

• Continued investments in sales and marketing have allowed Foundry Networks to capitalize on the early indicators of market traction Gartner observed a year ago. Revenue has increased substantially, and we continue to see this vendor mentioned in the shortlist discussions of a growing number of clients.
• Continued expansion of the workgroup switching line (to include more PoE options, including a cost-effective, full-power option) has allowed Foundry to offer a more complete product line. It is getting more traction for offering complete solutions, rather than being just a high-end LAN-core provider. This vendor is well-positioned for open convergence opportunities.
• Outstanding customer support remains a strong attribute that is repeatedly mentioned by Foundry customers.
• This is an extremely well-run organization, as demonstrated by sustained profitability and significant financial resources that ensure long-term viability.

• There is a lack of stacking in Foundry's workgroup products.
• This vendor's presence tends to be more geographically focused than other vendors; with its strong direct-touch model, enterprises need to ensure that they have the required local coverage from Foundry.

• HP ProCurve has a cost-effective, standards-compliant switch portfolio with strong edge features and broad PoE capabilities. Combined with its pioneering efforts around LLDP-MED, this puts ProCurve in a solid position to capitalize on the many open convergence opportunities in the market.
• A lifetime warranty and software upgrades across all workgroup products and the 8100/8200 core switches substantially reduce operating costs. This vendor's combination of aggressively priced hardware and lifetime warranty has radically changed the economic environment in the market.
• ProCurve has emerged as a significant force in LAN switching. Overall revenue for 2007 will approach $1 billion, and ProCurve has largely broken away from most of the alternative vendors as the No. 2 revenue producer in this market. Total port share in 2007 exceeded 10%.
• HP's major account executives get quota relief when selling ProCurve products, and we are seeing increasing support from the HP service organization. This increasing support from within the HP family will be a significant boost to ProCurve.

• There is a lack of a stackable offerings in the vendor's fix format products, which often forces the use of a 5400zl chassis product.
• A true enterprise data center product is still lacking, generally limiting ProCurve to enterprise connectivity solutions and small- to midsize-business data centers.
• ProCurve's free lifetime warranty often gets in the way of appropriate discussions regarding the complete range of post-sales services that organizations should consider.

• A strong overall portfolio and the introduction of the 45xx and 25xx series in 2007 drive full-stacking capabilities through Nortel's workgroup switch line.
• Nortel pioneered (and continues to have strong support for) high-availability, switch-clustering solutions through the use of SMLT and RSMLT.
• This vendor has shown an ability to sustain a level of execution during the past three quarters, which is a good sign for a vendor plagued by inconsistency. It has solidified its place as the No. 3 vendor (in relation to revenue) behind Cisco and HP ProCurve.

• Nortel has slow-adopting convergence standards. It is a good choice for a fully converged voice and data solution, and its extremely high level of PoE ports shipped indicates strong success in this strategy. However, work still needs to be done in standards implementation if Nortel is to be considered for open convergence opportunities, to increase its opportunities in the broader market.
• This vendor has nonexistent marketing in enterprise data. Despite a growing and sound portfolio and increasing revenue, Nortel rarely shows up on competitive shortlists.
• Nortel has spotty coverage in direct sales and channels.