Is it the end of the miracle for Brocade ?

Brocade was among the few pioneers who introduced, built and developed Fibre Channel product at the mid of the 90's. We all remembered Gadzoox, Chaparral, Vixel, a few others and Brocade who produced entry level switch product. The SAN era was launched, we heard at that time the word Virtualization, Lan-Free and Server-Free backup and a few other terms... promises were high and solutions delivered several good results for many years even today. Then Director wave arrived with big data center and consolidation projects, players such as CNT and Inrange was very well known and were considered as key reference vendors. Brocade tried to play in that space with no real success at the beginning, even if the brand was well adopted, recognized and synonymous of quality thanks to a large oem portfolio. Some people remembered the joke around the 12000 switch or director ? Then CNT bought Inrange for 190M$ in 2003, McData then acquired CNT for 235M$ in 2005 and the number of players was just 3 or 4 if we considered the HBA adjacent players such as Emulex and QLogic (QLogic acquired Ancor Communications in 2000 and tried to shake the palm in the switch/director space). Then an incredible fact occurred or I prefer to say not occurred, Cisco didn't decide to acquire McData and Brocade was able to absorb McData. It was a major factor to explain why Brocade still exists, just imagine if Cisco bought McData...

For a few years, iSCSI tried to penetrate the data center and the storage connectivity but the success was not as big as people thought. Ok but it works pretty well, who is behind iSCSI ? Cisco, Intel, Microsoft, IBM, Adaptec... and if they wish to really push that technology, no doubt about it, it will be a tsunami especially in SMB and server virtualization world. Many big enterprises trust and use iSCSI without any issues especially in server virtualization environment. We'll see for iSCSI but for sure, the perimeter of the data center is reduced, more density in term of port, speed and storage capacity, is it good ? I let you decide but who really need 8Gb/s FC for instance ? In the past, vendors tried to justify FC acquisition with aggregation of channels (trunking) to boost performance and now the reverse story is true: select a big pipe, divide it and dedicate some channels to virtual machines. This is virtual technology evolution.

For Brocade, the situation was tough, we remembered that Greg Reyes tried to approach Cisco in 2004 to sell Brocade for 9$ a share (http://www.theregister.co.uk/2006/02/09/reyes_brocade_sec/). We know the next steps. Mike Klayko, who arrived in 2003 with the Rhapsody Networks acquisition, understood pretty quickly that FC market will reduce and that technology can't do everything, this is in fact "just" a 2B$ market versus a 10-12 times (22-24B$) at least for Ethernet with different growth rate also. The gap will grow without any chance to be reduced and eliminated so Brocade had to consider some new area of growth, urgently. Foundry Networks was bought in 2008 for 2.6B$, 2 times the Brocade revenue at that time !!

Some other ideas came in mind with software effort with acquisition of Therion, WAFS agreements with Tacit Networks (acquired by Packeeter then by Blue Coat Systems) and then NuView for StorageX essentially. All of these were good ideas but so badly executed by Brocade. About remote offices, WAFS is dead as people realized that they need something more comprehensive to accelerate and optimize all data flow and protocols not only file oriented ones. WAN Acceleration and Optimization is the right approach and convergence of players validates this move. Back to NuView, Brocade acquired the company in 2006 for 60M$ and burned money at a pretty high rate. It was the idea to launch the FAN (File Area Network) concept, dead in the egg, to clone the SAN wave, 10 years earlier. The value proposition was obvious but that solution was consider as a "Nice to have" so potential users pushed back their budget for these solutions. And again some key failures in strategy, product developments and sales behaviors, how a so brilliant company can develop 2 products for the Windows CIFS environments and nothing for the Linux/Unix NFS one ? This is clearly a bad dream but they did it. EMEA region for many US companies is a fantastic opportunity of growth and Brocade File 

Business was a nightmare due to some sales directors and bad decisions. Brocade decided to fire all this group end of 2008 and beginning of 2009 and some people succeeded to protect themselves even with bad track record. And recent news are funny too. Rahul Mehta, CEO and Founder of NuView, sold to Brocade in 2006, came back from a semi retired period to take back control of his baby, the file storage software StorageX, he wished to "rebuy" the asset from Brocade. He was chocked like the industry, analysts and journalists that Brocade failed to deliver at least the same result as NuView did. Brocade refused, end of story.

An other tentative was the HBA with the acquisition of SilverBack Systems in 2007 to compete in that market and provide end-to-end connectivity and advanced data services such as encryption, virtualization... The 2 others gorillas in that segment are well established players and Brocade partners: QLogic and Emulex. Now with FCoE, the idea to build converged network adapter - the famous CNA - is a new promise to the market and the 3 players tried to solve that challenge quickly at a very attractive price.

For a few years, a new acronym - FCoE - tried to build and redesign the storage side of the data center. FCoE means Fibre Channel over Ethernet and really integrated the consideration of a lossless Ethernet layer. The game is over, Ethernet wins, Ethernet is everywhere, approved and largely adopted by all of us and vendors and various industries. We're now in the realm of commodities and Cloud started to redefine every IT shop. But who really need FCoE ? users running FC today !! and just these ones, the other will consider alternatives at lower cost especially as some vendors already published 40Gb and 100Gb Ethernet roadmaps. Again the idea of convergence is a reality.

About a few IT trends, it's key to consider SSD (Solid State Disk/Drive) and Cloud Computing/Storage. The impact of SSD is clear, architectures come back to DAS - Direct Attached Storage - model so the advantage of SAN is reduced not to say eliminated. It's not a question of price as ILM and Storage/Data Tiering allow to consider SSD for active data which is 10 to 20% of total data. The rest can be easily moved and migrated to capacity storage such as SATA connected with iSCSI, potentially FC, but why ? or secondary NAS for files. So globally this new model is very attractive, many new players arrive and promise new approaches. Cloud also contributes to a big shift as users 

consider more and more external IT services in a SaaS model. Just to illustrate, it exists more than 1600 online backup providers on the planet. So the challenge is more to select the right solution and provider. Also as a second example, it's easier, faster and economic to consider a cloud archive vendor with clear SLAs and penalties in the contract instead of consuming time, human effort and money for selecting, evaluating and deploying an on-premise solutions. And everyone can imagine maintenance cost, technology and data migration due to product obsolescence... This cloud will have an important impact on FC and SAN market, definitively.

With Cisco UCS announcement, Brocade with Foundry was selected by IBM and a few others. Again, a positive consequence not really created by the vendor itself. Last agreements - Motorola, Oracle, Thales, Blue Coat... - try to sustain market value and opportunity and justify the acquisition price between 4 to 6B$, and it's already high. In term of finance, each FY 2009 quarter revenue reach approximately half a billion $ and the annual goal - the famous 2B$ barrier - should be approached and touched.

In that storage segment, Brocade was the leader and almost the only player so the exit strategy was not so easy to determine. The company had to come back in a more competitive zone to find new area of growth and to be considered as a target acquisition. Who is able to pay for Brocade ? and Why is it interesting to buy Brocade ?

About the Why, here are a few ideas:

- Brocade is clearly the market leader and the obvious choice for Data Center Storage connectivity.

- Since Foundry Networks acquisition, the portfolio is really attractive for Ethernet, IP, wired and wireless, DC and remote office in addition to storage offering.

- The merger between Brocade and Foundry and the potential new one is a good information for users as vendor consolidation has a good impact on financial accounts.

- Same ideas about patent and IP controlled by Brocade today.

- Very huge installed base in both Data and Storage networking segments.

- Strong oem agreements.

- Future FCoE and CNA potential adoption to reflect an end-to-end strategy.

The Who is pretty easy except if some surprise will occur. No software companies will do the move, no reasons for that. The idea of Oracle was amazing, why a top layer application vendor needs plumbery ? Brocade is too deep for an ISV. Also, EMC is not the choice as they prefer Cisco and they sustain joint initiatives around VMware, Cloud and IT services. IBM and HP will reduce the value of Brocade as current oem agreements will be cut. Especially HP has a very good product line with ProCurve. Huawei, not sure at all it will be a surprise but what about Juniper, probably the most credible assumption with the ability to maintain all Brocade oems agreements and capitalize on past Brocade 

and Foundry success. We'll see, it's been a hot topic every year for a few years...

This author consults with leading institutions through GLG