Cisco and EMC: The Back Story

My youngest teenage daughter is starting to enjoy reading those celebrity tabloids: who's together now, who's breaking up, etc. 

Not exactly my thing, but - hey - it works for her.
 
Our tech industry is not immune from the same kind of speculation.  With both Cisco and EMC both playing prominent roles in the broader transformation of the IT industry, it seems that we're both in the rotation once again.
 
As part of my customer meetings, I'm now occasionally pinged by people who are curious as to my personal take on all of this.  Like anyone, I can speculate on the future, but the past is largely a matter of fact.
 
People tend to forget all the great stuff the two companies have done together in the past -- for our mutual customers. 

So I remind them.
 
And while no one can predict what the future might bring, the past is always worth considering ...

The Very Early Years
 
Remember, I've been at EMC a very long time (almost 18 years!), and I can clearly remember a time many years ago when EMC and Cisco were like ships passing in the night. 

We did our storage stuff, they did their networking stuff, we had many of the same customers, but no meaningful formal relationship.
 
The first inkling that our paths were starting to cross came in the mid 1990s when EMC was selling a boatload of SRDF remote replication, which -- of course -- usually demanded a non-trivial network technology engagement. 

We were somewhat short of relevant skills at the time, but we found that our friends at Cisco had this long-distance enterprise networking thing nailed: products, expertise, ecosystem, trusted brand with our customers, etc. etc.
 
We briefly toyed with the idea of reselling Cisco's products and expertise, but we realized we didn't add enough value to the equation at the time.  Better to have Cisco and their partners do what they do best.  EMC did a long list qualifications, reference architectures to support the end solution -- all good work.
 
As the 1990s drew to a close (and the dotcom market started to bubble), we all started to realize that we could do better when we approached customers together.  If your memory is good, you'll remember the ECOstructure alliance -- EMC, Cisco and Oracle -- where we positioned our jointly-developed reference architectures as a good platform for building a scalable internet business.
 
The ECOstructure alliance was fabulously successful; the dotcom bubble somewhat less so. 
 
But we at EMC did come away from the experience with a very strong appreciation of what it meant to work closely with world-class partners around shared customer requirements. 

We knew we'd be doing more together.
 
The Storage Years -- You Want To Do What??
 
Let's move on to early 2002 or so -- and all us IT vendors were starting to slowly recover from our post-bubble hangovers.
 
In the EMC storage world, the majority of enterprise storage was SAN attached: neither NAS nor iSCSI were particularly mainstream at the time. 

If you remember, at the time the enterprise SAN switch/director market was largely a duopoly, fiercely contested between Brocade and McData. 
 
EMC had long ago decided to invest heavily in both partnerships, and – as a result – had created a significant amount of value-add on top of those core products.  In that era, EMC was responsible for a significant chunk of the SAN fabric revenue in the industry -- still true today.
 
Cisco had been watching this SAN datacenter opportunity, and -- for all the right reasons -- wanted to play.
 
If you're not familiar, Cisco has this fascinating spin-out, spin-in model for getting into new markets that has turned out to work quite well for them. 
 
A bunch of really smart Cisco people go off and set themselves up as a "start up" (with Cisco being the major investor), they build something really cool, and it then gets folded back into mainstream Cisco when the time is right.
 
In this early incarnation of this model, a small group of ex-Cisco engineers had set themselves up as "Andiamo" and were looking to build a next-gen platform for enterprise SAN.  They came to EMC with an interesting proposition: they thought they could build a far better SAN fabric technology base, would EMC be interested?
 
Now, remember at the time, EMC had a robust and thriving SAN business, so there were those here who were quite skeptical that the world needed a better mousetrap from Cisco, especially as we were doing quite well with our existing SAN partners at the time. 
 
But I remember the Andiamo team was quite persistent, and kept pointing to all the things that could be done better in the SAN. 
 
Part of that was bringing enterprise network management viewpoints into the technology model; that much was obvious.  The other cool idea at the time was the notion of programmable intelligence on the switch itself which could be useful for such things as storage virtualization.  In hindsight, this second idea didn't really pan out, but I remember it being pretty exciting stuff at the time.
 
Before long, EMC and Cisco ended up structuring a formal relationship to help get Cisco into the enterprise SAN market (through EMC) with what ended up being the wildly successful MDS product line. 
 
Remember, this wasn't some simple resell arrangement; EMC had to build an entire business around the new joint offering. 
 
This entailed an enormous amount of work by EMC: extended engineering qualification work, integrating our storage management tools, building a new supply chain and support model that worked well with Cisco's capabilities, training our presales, sales and partner organizations, managing the now-shifting relationship with our existing partners Brocade and McData, and so on.

This also entailed an enormous amount of work by Cisco: they didn't have a model for this business: the technology, how it flowed to market, how to engage with the customer before, during and after the sale, how to work with the established partners, and so on.  Much heavy lifting on their part, with some coaching from EMC.
 
Both Cisco and EMC signed up to make MDS a huge success, and the rest is history. 
 
The real winners here were our mutual customers, I think. 
 
First, they had access to some unique and differentiated technology from Cisco, all supported and integrated by EMC.  Not only FC, but data-center-class IP for not only cutting-edge NAS and iSCSI implementations, but long-distance replication as well.  And, of course, without the MDS, there never would have been a Nexus.  All good.
 
More importantly, the success we both saw helped fuel mutual interest in newer technologies, such as FCoE, and -- yes -- the Cisco UCS.
 
The Server Years -- Cisco Does Servers?
 
Fast forward to 2007.  Many of the same players we had worked with on Andiamo were now spun-out again -- this time dubbed "Nuova" -- and looking at the server marketplace.
 
They believed that -- by converging servers and networks together -- they could do a far better job on server resource and management efficiency. 
 
It didn't take long for EMC to appreciate that they had a very good point, and -- by bringing Cisco's network DNA to the server world -- it was clear they could create a very differentiated and attractive alternative to what was already in the market from HP, IBM and Dell.
 
This time around, though, there was a lot less resistance internally at EMC.  We were entirely smitten with the notion of an "array of networked blades". 

The EMC enthusiasm around "Project California" got to be distracting to other, more pragmatic things we had to get done.  Early on, it was clear we wanted to everything we could to make UCS a success.
 
Once again, we toyed with the notion of reselling the Cisco product, but we couldn't see it really working out, given the compressed nature of margins in the server business. 
 
But we could do an awful lot to help make the UCS successful: extensive qualification work, being able to provide first (and second) level UCS support through EMC customer services, integrating our management tools, application profiling for things like SPA, incentivizing our sales teams to introduce Cisco to the parts of the data center we as EMC were more familiar with, and so on.

In some sense, EMC was a broad enterprise IT company without a server to sell -- we had all the other pieces, though!  And Cisco's new offering fit nicely in from our perspective.

Just to be clear, we never believed that Cisco would have an exclusive relationship with EMC around storage, nor would we at EMC believe we would ever have an exclusive relationship with Cisco around servers. 

That's not how the world works. 
 
But we saw a clear opportunity to do more for our mutual customers, especially as Cisco was getting into an important new market.  I clearly remember many dozens of customer discussions at the time when they wanted to know what EMC was doing with the UCS, often when Cisco wasn't in the room :)
 
We told it like we saw it: UCS was very cool and differentiated server technology.  We at EMC work closely with all the different server vendors, and we believed the Cisco approach had the potential to be the best by a wide margin.  Cisco was clearly committed, and so were we.
 
As a result, we at EMC were going to invest serious amounts of money and resource to make sure UCS works first and best on EMC products, because we though that was what many of our customers will end up wanting.  Backed up by plenty of mind-numbing detail, if you were interested.
 
I think that clear and unwavering message of EMC support gave more than a few customers the courage to invest early in UCS when it was comparatively new and unfamiliar.   Over time, though, confidence was earned, and Cisco continues to innovate and differentiate in the server space.
 
Once again, I think our mutual customers won big.
 
The Solution Years -- Is There A Better Model?
 
Once EMC was committed to making UCS a success, and the plans were locked down, the discussion took a decidedly different turn. 
 
Howard Elias, in particular, latched on to a simple yet powerful idea -- was there a better way of getting all these technologies to our mutual customers?
 
You could see where VMware was going with virtualization and new operational and management models.  You could see where Cisco was going with UCS -- an architecture optimized around virtualization and new operational and management models. 

And, of course, on the storage front EMC was doing many of the same things.
 
Rather than put each vendors’ catalog of separate ingredients in front of our mutual customers -- since we often end up talking to the same sorts of customers -- could we bake an easy-to-consume meal for them?
 
Howard wanted to try, and pushed a strong agenda between the two companies.  The first iteration of this joint approach -- Acadia -- had many of the right intentions, but wasn't really organized for success.  

Acadia was basically an alliance with reference architectures, not a product.  Nice, but it was still largely a roll-your-own approach for customers, albeit incrementally a bit easier. 

Bottom line: we really hadn't changed the game the way we thought we could.
 
I'm going to give Howard Elias full credit for then proposing a more radical approach: create a separate joint venture with the express mission of making converged infrastructure easier to consume and easier to operate. 
 
The result was the now-familiar VCE and Vblock -- a model whose widespread success and value proposition still hasn't been reverse-engineered by the competition. 

Both EMC and Cisco contributed a non-trivial amount of technology, money, resources and talent to VCE -- a substantial investment with a correspondingly big payoff.
 
Once again, our mutual customers won big. 

They now have access to a completely different infrastructure consumption and operational model -- whether they did it internally, or using any number of compatible service providers, or -- more frequently -- a combination of the two. 

Together, we ended up changing the game.
 
A few years into the exercise, we all realized that we needed a subset of what VCE was trying to do with Vblocks, but in smaller configurations, with more technology choices, and an increased focus on partner value-add. 
 
The result was the now-successful VSPEX from EMC, a solution-led approach where Cisco technologies and CVDs (Cisco Verified Designs) play a key role.
 
Once again, our mutual customers (and partners) benefit.
 
There's Much More, But ...
 
If we were to dig deeper, you’d find even more we could talk about.
 
For example, the two companies have done a ton of work around security (mostly by integrating RSA capabilities with Cisco capabilities), virtualization (VMware and Cisco working together on products like the Nexus 1k), application-specific solutioneering (VDI, SAP's HANA to name a few), and much more.
 
Even in the infrastructure space, combining Cisco's OTV and EMC's VPLEX in the form of a Vblock creates a fascinating set of workload mobility capabilities that have to be seen to be believed.
 
From my perspective, the details aren't all that important; the broad picture is -- both companies now have a deep and familiar pattern of investing together to create better solutions for our mutual customers.  Not in one or two opportuntistic places, but in as many places as we can find meaningful engagement.
 
And we're always on the lookout for more.
 
The Road Ahead?
 
I can't speak officially for either EMC (or Cisco!) in this regard, but -- from where I sit -- there's plenty of cool potential on the horizon that we might decide tackle together. 
 
For example, both companies are very interested in mobility and collaboration; there's a few interesting joint opportunities there.  Of course, we're looking forward to Cisco's Insieme project (yet another spin out!) down the road.

Our mutual interests in the intersection of big data and video goes without saying -- analytics from video streams, anyone?   In the future, big data will inevitably have to be federated across multiple geographies -- it's just too big -- which puts Cisco's capabilities once more in the spotlight.
 
And much, much more.
 
Once again, we'll see amazing opportunities for our mutual customers and partners that inevitably result when you mash up one incredibly broad technology portfolio with another. 
 
Will everything we try together work out?  No -- that hasn't happened in the past, and it certainly won't happen in the future.  I can point to exciting things we've tried together that haven’t panned out for one reason or another.  But we keep trying.
 
Will we occasionally have differences of opinion over one thing or another?  Yep.  It was true 15 years ago when we started working together, it's true today, and it will certainly be true in the future. 
 
That's what I believe makes a good alliance work well -- that differing perspective.  Otherwise, there wouldn't be much of an incentive to work together, right?
 
But the real reason we work so closely together?
 
Our mutual customers expect us to.

Comments

[Giovanni Cintorrino]

Great post Chuck. I never knew Cisco was such an Italian company...with project names like "Andiamo" and "Insieme" being spun up.

[DaniC]

Very interesting post, well done!

Interesting to see how 2 giants work together...