Peter Christy

How Microsoft Lost the Virtualization War (when cheaper is wrong ...)

Microsoft didn't have to lose data center virtualization to VMware. It wasn't a question of inferior technology or product functionality. They lost because their business model was wrong.

Microsoft's Server and Tool Business (STB) has done extremely well by identifying markets where complex software functionality can be engineered for simple installation and use, and then brought to market at a disruptively low price point. Time after time this model worked: Windows Server, Systems Center, Exchange and SharePoint. Not surprisingly, when it came to virtualization, this was the business model: offer lots of functionality at a disruptively low-price. Why did it fail so badly in this case?

It failed because virtualization was different. Virtualizing a set of servers isn't like upgrading the operating system or Office. Virtualization enables significantly improved economics but along the way it requires significant changes in how the data center infrastructure is managed and how the data center is organized. Therefore it isn't surprising that customers didn't just jump into virtualization. Instead they brought in consultants to help them plan and implement, and along the way created a new channel focused on the delivery of virtualization solutions.

VMware had an "expensive" solution (although still a very smart investment considered by itself). Selling the VMware solution made a lot more sense for the channel because they made a lot more money doing it, and at beginning the VMware solution was domonstrably "better" than the Microsoft solution (more advanced functionality). So if a channel partner could make a lot more money selling VMware guess what they did?

As the Hyper-V solution caught up (and with System Center in many ways surpassed VMware) guess what happened? More or less nothing, because the channel had no incentive to switch brands. Yes, VMware has an "expensive" solution, and customers would like a "cheaper" solution, but that's not enough given the partner economics.

I finally understand what people were trying to say when they noted that Paul Maritz knew the Microsoft play book ("he had written it").  The genius wasn't just to understand your big competitor well; it was to jujitsu Microsoft (use their own weight against them).  The bigger question is why Microsoft didn't see it coming. The ultimate loss is enormous. Windows Server has done extremely well in the data center. System Center is remarkable management system. But now flanked by VMware on both sides it's going to be challenging to get out of the pincer.  Ballmer has changed the STB management. Microsoft is a smart and powerful company. Next move is necessarily theirs.

A Shoutout to Marvin Wheeler ....

I just learned that Marvin Wheeler is leaving Terremark, having completed their transition into Verizon. Marvin is one of nicest and most interesting people I've worked with.

I met Marvin at EMC's "Road to the Private Cloud" analyst meeting late in 2009. Terremark was a featured EMC customer "poster boy" talking about the government business they had been capturing into their new Cloud data center offering. I asked Marvin what the key to their value proposition was. He said that his answer would probably surprise me: it was the security that they provided. He was right. At the time (and to a large degree still today) security shows up on the "con" side of the Cloud ledger, not on the "pro." I asked Marvin if we could follow up on the topic and he arranged a follow on briefing with Chris Day, Terremark's security leader.

Marvin explained that five years prior he had hired Chris, noting that whenever computer crime was mentioned in the Miami papers Chris was always the expert quoted. In that time, Chris had built up an interesting set of security services sold togovernment clients and in support of areas like forensic accounting. They specialized in what you did after you had been penetrated, not just preventing penetration (until recently, the prime focus of the "security companies"): discovering penetration; understanding the nature of the threat and scope; stoping the threat; remediating the damage. Terremark had realized that in a virtualized SP data center they could offer versions of these services that were pre-wired into the infrastructure fabric (rather than having to be patched in after an attack) and they could provide the expertise to use these offerings whereas their government customers couldn't on their own. On balance, although you might be slightly more vulnerable to penetration in a multi-tenant data center (or at least it was arguable), you were clearly better off if you were penetrated. Terremark was way head of the rest of the world being able to view the Cloud as a new medium with new opportunities rather than something where you did what you always did.

I was impressed and added Marvin to my "bright people worth keeping in touch with" list, and discovered along the way he was one of the "nice" guys in the business, always possible to reach and always useful to talk to.  More recently, Marvin as been central to the formation of the Open Data Center Alliance that was spun out of CIO discussions that Intel had.  I would guess that the Cloud purists think this is another example of how hopeless private clouds are, but ODCA is an important venue for making progress on building practical hybrid Cloud integrations.

I can understand completely why Terremark all of a sudden getting much bigger quickly is a time to move on. I'm sure Marvin will show up somewhere interesting soon enough, and wish him the best!

VMware VSphere 5: So What?

Last week VMware introduced vSphere 5 with a very professional, Web-based launch (best Paul Maritz talk I've ever heard). The reaction has been minimal... Why?  I think there are two reasons:

1. VMware doesn't really have any competition any more.
2. The adoption of Cloud solutions in the Enterprise is limited more by the rate at which the customers can evolve than by the state of the product offering (VMware is nicely ahead of the customers).

Two years ago, my thesis was that Microsoft would catch up with VMware in terms of virtualization functionality, and then bring the significant assets of System Center to bear competitively. But that just hasn't happened. Microsoft talks proudly of their share gain in hypervisors, but my friends at IDC say that really isn't gain against VMware in the data center. System Center is still a fine offering but Microsoft has nothing like vSphere as a platform offering on which to build private clouds (as you know, where I think the business is today). Microsoft is sandwiched between a VMware hypervisor and VMware service management. If that flanking continues, I don't Microsoft will thrive. 

This week both VMware and Microsoft announced earnings. VMware announced a very healthy 40% Y-o-Y growth rate. It's harder to parse out exactly what System Center revenue growth was but my best guess is that it's something in the 15+% range, not bad, but certainly not keeping up with VMW. A couple of years ago Microsoft argued that VMware and System Center were the only parts of the systems management segment that were growing, and that each was growing about the same rate. As you can see from the chart before, for a while VMware wasn't really growing license revenues: the overall growth came from a increasing service line with a growing installed base. But license revenue growth is back up.

 
Maybe there is another way to view the data, but it looks to me like VMware has sort of won this battle over the last year.

Cisco: And Now the Wake

Not unexpectedly, Cisco announced their immanent intent to lay off roughly 4,000 people in addition to those who took early retirement, as part of making whole a $1B expense reduction, as well as the commitment to streamline the company functioning.

Although the layoffs were not discussed in any of the general Live! sessions, John answered a lot of related questions in the Analyst meeting Q&A session. Not surprisingly, he's good at painting the positive side of the story (more efficient and responsive company). I tried to get him to explain what this did to the revenue growth plan for the company going forward (remember that original intent of the boards and councils were to accelerate and democratize innovation). He wouldn't/didn't reply to that directly, but instead focused on the fact that when growth is lower you can't spend as much on innovation (true, but leaves open the ultimate question of revenue growth and market valuation).

He suggested that the plans going forward were still in progress, so hopefully a deeper discussion is forthcoming.

One of the tantalizing aspects of the promised layoffs (that will be "operationalized" in the first week of August) is that the ax will come down more on people with VP and higher title (15% decimation). Now we just wait and see who ends up in the sights..

Cisco Live! Wake or Celebration?

I've never been to a Live! meeting before. I'm fascinated by how each user group has it's own personality. Live is a lot more like Microsoft TechEd than DreamForce: largely male geeks, look to be in somewhat better shape than programmers. Whatever worries the industry has about Cisco, they weren't apparent here. Cisco blew out the attendance with a 30% growth over last year (15,000). No scrimping was evident at the show with a big and loud exhibition and dancing girls warming up the keynote crowd (which began, by the way, with a Flash Mob including dancers dressed as attendees and ushers).

Chambers seemed somewhat subdued in the keynote although a subdued Chambers is still very animated by any other standard. The only real news in the keynote content was the elevation of security to a first priority. Video of course is still important; the Flip was not mentioned. Separately Cisco announced a major refresh to the Catalyst 6500 (first shipped in 1999 and remarkably representing $43B of revenue). The refresh doesn't seem a retreat from the Nexus 7000 as much as a way of offering a lower-performance, lower-price option to existing customers. I asked people about "Jawbreaker" (the rumored new data center offering) but no one was talking although there was some "wink-wink, nod-nod, do you know what I mean."

What's still not clear yet is what Cisco's business strategy is going forward, net of layoffs and shedding of businesses. Chambers is doing a Q&A with the industry analysts later this morning but I'm not holding my breath.

Citrix Acquires Cloud.com -- What's the Market?

Today Citrix announced the acquisition of Cloud.com.  Cloud.com is clearly a leader at providing an alternative to VCloud, and one based on Open Source and standards. They aspire to be the "Red Hat" of OpenStack.  From this perspective, they fit will into the Citrix Cloud initiatives. But all that said and done, I think it's worth asking the more difficult question about the market. Who buys Cloud platforms and why?

Some years ago, Microsoft argued that there was only a need for a small number of Cloud fabrics because of the economies and market value of scale. Others argued that there would be tens of thousands of Cloud SP's, each serving some valuable niche based on platform (Iaas, PaaS, SaaS), locality and vertical. The later argument is the basis of the energy in the market. Clearly if Microsoft is literally right then there ultimately isn't much of a market for something like Cloud.com.

To try and understand this interesting question better, we've actually talked to a number of SP's. We found some SP's that hadn't thought about the nuances much but also quite a few that had. The knowledgeable ones didn't think that competing with Amazon for what AWS does well was an interesting business opportunity, viewing it instead as a "race to the bottom." They were pretty uniformly focusing insted on some more specific offering, often based on a higher level platform (e.g., supporting SaaS offerings) rather than offering rented VM's.

Now here is where it gets interesting. Many of these more sophisticated SP's also wondered whether over time it would make more sense to build those offerings on something like AWS or Azure rather than having their own under pining fabric. So it's possible that in an important sense Microsoft might be right. The world may only need a relatively few fabrics, just like we need a relatively few power grids or telephony systems. At the same time, those who argued that 10,000 SP's would bloom might be right too, presuming they build on those massive fabrics.

Citrix didn't disclose the price of the acquisition so it's fair to assume it wasn't literally a big deal since it isn't material to Citrix stockholders. If that's true, the industry also may aslo believe that ultimately the market for Cloud.com was limited.

 

It Should Be Quite a Week ......

Next week (July 8 - 12) should be exciting (hardly the summer doldrums). I'll be at Cisco Live in the fast part for their annual analyst event. It's going to be illuminating to hear what the story is now and how it's changed since they hit rough waters. In addition to problems in the core switching business, a lot of the organizational changes recently were motivated to fix the problems done with the goal of optimizing  innovation. So there is no easy answer: the core business has issues, and innovation is more important than ever.

Then, on Tuesday, VMware is making a major announcement about the next release of VCloud. Last time, the new additions in security came at some cost to their ersatz security partners. Presuming that this release incorporates additional "adjacent" functionality, it will be fascinating to see the new implications for the partner ecology. And you have to be curious about why the announcement is now rather than capping the upcoming VMworld meeting.

And on Monday Citrix is going to be talking (at least to the analyst community) about the Cloud. You have to wonder whether this is to keep from appearing reactive to whatever VMware is going to say.

Anyway, it should be an interesting week!

Open Source and The Cloud -- The Conundrum(s)

Open Source has always fascinated me. On the one hand it make so much sense; on the other hand it requires imagining a conversation between Karl Mark and John Doerr. Up until recently at least the economic model was understandable:

• Open Source offered free functionality for individuals or students who couldn't afford it otherwise (especially in a modifiable or extensible form).
• It was also free to the largest users (like Amazon using Xen), since they had to support it themselves anyway.
• There was a business to be made supporting the users in middle who were too small to support it themselves but too dependent on correct operation to responsibly go without support.
• It was also possible to add your own proprietary and separate modules, and sell them as separately licensed software (the rules say that improvements to the base software have to be contributed back to the community).

 A recent encounter with old friend Chris Keene (most recently CEO of WaveMaker, now a strategist at acquirer VMware) added new facets to fhis problem (see his earlier post).

• There is a loophole in the basic GPL license that says that if you use Open Source to deliver a service, but you don't sell or distribute the software, you don't have to contribute your changes back to the community.
• When Open Source functionality is delivered as part of a Cloud service (rather than as software), the support proposition vanishes -- you get reliable use in the form of a service but you don't have to maintain it yourself. So as more use moves to the Cloud, and is delivered as a service, the established Open Source economy crumbles.

Ah, but things get stranger still if you think that OpenStack and Hortonworks are leading indicators. Before OpenStack, NASA was using Eucalyptus which consisted of the Open Source bits and the proprietary bits. NASA wanted the proprietary functionality as part of the Open Source base and when that negotiation failed they partnered with RackSpace and just did it themselves. Hortonworks is taking the same approach: they think the management functions that Cloudera adds to Hadoop are valuable base functions so they are going to add that functionality right into base distribution and then give it back for free to the community.

So the big players, using Open Source, have special priviledges (not required to contribute additions back to the community) and can also destroy the part of the Open Source business model that leverages proprietary additional software.

Hmmm... Sounds to me like Karl Marx (or maybe Stalin) 1, John Doerr 0, at the end of this inning.

 

Yes the Cloud Will Cost IT Jobs

One of the topics that was gracefully skirted at Structure 2011 was the question of whether Cloud computing would cost jobs in IT.

The honest answer is of course jobs will be lost:

• Remember that IT is a cost center and cost savings is always good.
• Remember that Cloud technology (public or private) brings automation that replaces manual activity.
• Remember that the use of public Cloud offerings enable a trade between fixed and variable expenses that includes the use of permanent staff.

So it should come as no surprise that the increased use of Cloud technology and architecture will diminish the ranks of some the traditional admins in a data center. The story is by no means bleak for those who work in enterprise IT. The Cloud impact also includes a general up-leveling of people in the data center (e.g., a need for system admins who can understand, for example, application performance in a more complex and shared environment). And, as Randy Mott tells the HP story (Randy Mott's Secret Sauce), he was able to diminish overall IT spending considerably while at the same time increasing the amount spent on innovation and forward-looking activity. So less can actually be quite a bit more.....

 

GoDaddy Gets Very Deep Pockets (the data center business gets red hot)

GoDaddy just announced that they have taken an investment from and formed a strategic relationship with three of the largest private equity firms (KKR, Silver Lake and TCV). The financial details weren't disclosed but it's fair to say that a company that has grown largely by self-funding just got access to a very large war chest. This follows closely on the finalization of Verizon's acquisition of Terremark and the integration of the organizations, in essence quickly bulking up Terremark and giving them access to the resources of Verizon.

Now I guess we just have to wait and see exactly what money buys....

Cloud Opportunities: The Willie Sutton Thesis

My "Big fat idiot" post has lead to lot of interesting discussions on why enterprise IT (the "Private Cloud") should or should not be considered to be "Cloud." My logic might be wrong, but couldn't be simpler: Just like what Willie Sutton said when asked why he robbed banks, it's "because that's where the money is."

When we work with clients to help them understand emerging market opportunities, our ultimate focus is always to follow the money and understand how a new product or service delivers tangible value to the customer (i.e., is a good investment), and from that perspective, to help them understand in which vertical or use case the aggregate value is greatest. That's how I choose to look at Cloud opportunities as well: where's the money to be made?

In an earlier post (Why Clouds Exist) I argued that we've had a remarkable confluence of events without which neither Public nor Private Clouds would exist today, including these specific contributors:

• The consolidation of applications into data centers that started 10 years ago.
• The victory of Intel's X86 architecture.
• Virtualization enabling mixed platform X86 workloads to be run on a shared server.

But the Cloud isn't about virtualization; rather it's about how virtualization enables long-standing visions. In the Enterprise that means one thing:

• On demand computing: the ability to automate data center operations (take manual activities out of the provisioning loop). All of a sudden the management of many specific applications becomes much simpler, the application quality improves, and the cost of operations goes down.
• Service-Oriented IT: Automate what IT used to do (e.g., buying and racking a server; provisioning the O/S and software load). Automate everything possible. Let the business drive use. Focus on forward-looking enablement. By making this transition, IT becomes a more important part of the business.

Outside of the Enterprise the Cloud has very different benefits:

• Infrastructure as a service: AWS and a credit card eliminates the need to buy and operate computers. This takes capital spend out of new venture development. This is unquestionably a revolutionary change and utterly transforms how systems can be funded, built and operated.
• IaaS in turn enables whole new IT ecologies. iPhone or iPad applications with some server requirement would never have happened without the ability of App developers to exploit AWS for the backend.
• From this perspective, it's also true that applications designed for this new world have very different architectures (stateless, RESTful, ...) that is very unlike legacy application architecture.

I agree with my detractors that IaaS on AWS has very little to do with what is called the Road to the Private Cloud. But that's not my point. It seems to me that Willie Sutton would have chosen the Private Cloud because it seems to me that's where more of the money is. The Private Cloud is as important potentially to CIO's as AWS is to startup, albeit for entirely different reasons. But they are going to spend a boatload of money because that investment will save them even more money!

 

The Hadoop Summit -- The Plot Thickens

The Hadoop Summit this week was a bang up success by any simple measure (like attendance or enthusiasm). The cornerstone announcement was the formation of Hortonworks by Yahoo and Benchmark Capital, spinning out key Yahoo Hadoop resources. Like all Hadoop commercial efforts there are some delightful twists (Cloudera reasonable suggested some clarifications below):

• Like Cloudera, Hortonworks will offer Hadoop support services.
• Unlike Cloudera, Hortonworks will contribute all their developed code back into the Apache Hadoop distribution project (Cloudera has their own distribution; Cloudera contributes all Apache Hadoop changes back to Apache and develops proprietary management modules as well ).
• Very unlike Cloudera, Hortonworks will specifically invest in the management capabilities that is the "secret sauce" by which Cloudera differentiates the not-free Cloudera Enterprise Edition and contribute these additions back to the Apache project.
• For Cloudera, Hortonworks represents incremental validation of the market (good news). But Hortonworks sure smells like a serious competitor to me.

For me the critical business question is how to make money with Hadoop (clearly as a free tool Apache Hadoop provides great value). Since most of the Hadoop dialog is the enthusiasm, I've always felt like a cranky voice and wet blanket,even though how to make money is something everyone in the community should be concerned about because ultimately it is that revenue stream that will really define what Hadoop becomes. So I was just delighted by IBM's Anant Jhingran's keynote (CTO, Information Management) because Anant spoke directly to exactly these issues.  I strongly encourage everyone who wasn't there to watch when the video goes up.

Anant painted what to me is a remarkably insightful view of Hadoop and how it fits into the existing world of analytics, and from that perspective how to better understand the business issues. IBM clearly has an important analytics business already, so it's not surprising that Anant doesn't say that SQL is dead and that Hadoop and nonSQL will take over the world. So you could cynically say it's just legacy company FUD. I think not: I think it's the wisdom of someone with a lot of experience in the business.

Some of his key points (to me) include these:

• Don't get hung up on "big".  Hadoop was key to IBM's Watson solution but after the set-up the core database was just 6 TB which isn't that big ("volume" isn't everything).
• Remember that most analytic solutions use a diversity of data ("variety is important").
• Remember that for some solutions the ability to ingest rapid streams of data is an important element ("velocity").  He showed a slide in which IBM had ranked various important analytics solution are decomposed by the degree to which Variety, Volume and Velocity are important. Anant didn't take the next step and describe the role Hadoop should and can plan, but it certainly seems like a valuable framework for considering that problem.
• He colorfully described the Hadoop community as consisting of "birthers" (those on the leading edge, pushing the boundary with Hadoop) and "adopters" (the larger BI community that would like to integrate Hadoop into their existing and valuable solutions). He reasoned that these quite different groups needed to find a way of working together and working to the broad common good (rather than splintering into factions).
• Anant also pointed out the hidden obvious truth (the ignored elephant in the tent if you forgive the egregious pun): it's really hard to make money from Open Source.

Like many said in one way or another, 2011 is clearly a watershed time for Hadoop and Big Data. A lot of the complexity in the paths to be taken and the decision to be made surfaced at the Hadoop Summit. Now to just consult my crystal ball....

What's a Cloud? (Simon Crosby is a big, fat idiot)

OK, I admit that's cheap sensationalism. Simon Crosby is not "big" (tall, yes) and he is anything but fat or stupid. But I think that he gets a couple of things seriously wrong when he defines to Cloud to only apply to services you can rent by the hour, and when he rejects all of the Enterprise IT efforts as "Cloud" relevant because it is too labor intensive.

Cloud is an very relevant topic because of the dramatic and ongoing change in data center technology and operation. This acceleration is unquestionably catalyzed by production quality virtualization but the change isn't about virtualization; it's about automated "on demand" and "elastic" operations and how they accelerate the practicality of business models like service-oriented IT.

Simon's exclusion of the Enterprise activity (EMC's "Road to the Private Cloud") is wrong simply because that is where the majority of IT Cloud spend is occurring, and because it meets our definition precisely. Simon is absolutely right to point out that enterprise IT cost structures are very different from Amazon or Google cost structures. Today CIO's yearn for a day when the cost of the hardware is a primary issue but today it's not. Private Cloud computing is a pathway toward a cost structure that is much leaner in labor, but getting to that point will take many years. What Randy Mott did at HP ( Randy Mott blog entry ). It's a painful process because it requires decommissioning old and redundant applications, changing business processes and retraining people, but it's all an important part of getting to service oriented IT. Don't interpret a very different beginning point as enterprise IT not having the same aspirations or not finding great value in Cloud systems.

Simon says that it's not Cloud computing if you don't buy it as a service. When AWS first got wide attention, more than one guru said that no one would do private computing again because AWS was so much cheaper. Now with the likes of Netflix, EBay (and even amazon.com) implementing hybrid public and private strategies, the real economics are becoming more visible and common sense is prevailing. AWS is a lot more expensive than what EBay does internally, but if you have a highly peaky workload, the economics of a hybrid scheme still make sense, even if AWS cost 4X your internal costs.

Here's the critical point that isn't broadly understood yet: normal enterprise workloads aren't highly peaky, and if you put enough of them together in a large, shared data center (and movethe peaky dev/test and marketing loads out to the Cloud where they rightfully belong), enterprise IT demand is highly predictable and remarkably constant. So an automated, on-demand, large enterprise data center can run at much higher utilization than can AWS (because AWS is providing peak capacity for many), and it doesn't have to make a profit. Enterprise Private Clouds can and will be as automated and nearly as labor efficient as AWS (at the infrastructure layer) and it will be significantly cheaper for enterprises to run their own private Cloud than to outsource to Amazon, not to mention the clear performance and governance benefits.

It his previous Citrix role Simon had many reasons to argue that what VMware is doing isn't really Cloud computing but if you look at how enterprise IT is spending money and what their goals are, it would be a tremendous mistake to exclude this from the Cloud discussion.

Why Clouds Exist

At times it can be hard to understand what part of the Cloud discussion has substance and what is just marketing fluff. What I want to do here is argue that over the last ten years there as been a "perfect storm" of events that collectively led to today's Cloud world. Clouds are not just yesterday's time sharing or ASP offerings with fresh marketing lipstick. Here's the story.

Consider all that has happened over the period time since the Dot Com bust:

• The Intel architecture "won."  If you were planning a Dot Com venture your funders would probably tell you to use an Oracle database on Sun Solaris. In 2000, many businesses had strategic investments in the IBM Power Architecture (or HP's proprietary architectures). Today there is widespread agreement that Intel's X86 is good enough. Intel won the microprocessor battle. Without that architecture convergence there would be no Cloud.
• The network got a lot better. Broadband access wasn't common ten years ago. It is now. Without modern connectivity and bandwidth there would be no Cloud.
• Wireless networks expanded. Wireless build-out during the period was remarkable on a global basis bringing developing areas immediately into the networked world, bypassing entirely the wired phase.
• Mobile devices became powerful and smart. With Apple's iPhone deal with AT&T, mobile applications escaped the control of the mobile carriers. The evolution in wireless and mobile devices (and the rapid rise of consumerism -- employee chosen devices) has really changed the enterprise application and design model. This transition is an ideal time to also transition the front-end parts of the application to a Cloud implementation.
• We consolidated server applications into data centers. During the Dot Com boom, businesses spent profligately for IT, fearing that otherwise they would be "Amazoned." When the bust came there was broad retrenchment in IT spending (it's a cost center, remember). One of the easiest ways to reduce IT cost is to consolidate diverse applications into data centers where the administration and support can be local and shared. This cost-driven consolidation was an important precursor to today's Cloud world.
• Google demonstrated the promise of very large scale commodity computing. Ten years ago if you suggested building "mission critical" applications on anything other than "enterprise class" servers and storage you would be laughed out of the room. Google changed the game entirely. They showed that the most powerful systems in the world could be built from high-volume, commodity parts, and operated with high availability at remarkably low cost. In part the drive to the Cloud is motivated by the radically improved cost structure clearly visible downstream.
• Intel adopted multicore microprocessor design.  For years Intel converted semiconductor design and process improvements (the train under "Moore's Law") into faster CPU's by making the CPU's more and more complex.  In the end complexity yields diminishing returns. Early in the decade progress looked threatened: new processors got the same task done more quickly, but required more energy in the process. Finally Intel followed AMD and adopted a multicore strategy in which technology improvements are used to put more independent processors on the chip (more "cores") rather than to build more complex processors. It was a risky decision. It wasn't clear how many cores a desktop PC could use productively and the same problem existed in the data center where Windows servers were commonly built to each run on an separate server. The adoption of multicore designs enables a rapid increase the power of a data center of a given size and with a given power budget.
• VMware created commercial quality X86 virtualization and server consolidation occurred. Virtualization abstracts a workload from the specific details of the hardware on which it runs and enabled the shared use of a virtualized server by many X86 workloads with no requirement that they all ran the same "stack."

Virtualization is the magic lubrication that brings it all together: we consolidated applications into data centers; the victory of the X86 architecture and virtualization enabled running many disparate workloads on a shared server; multicore enabled an uptick in performance and power efficency improvements; Google proved that large scale data centers could be built from commodity technology and operated with amazing economics. The rapid evolution on mobile access and devices are transforming the requirements for enterprise applications providing a high-value motivation to change application front end architecture, just at the right time to leverage Cloud implementations.

There is hype and fluff in the Cloud discussion but underneath that there is a remarkable confluence of independent technology waves. In this case the substance is far greater than the hype.

The Cloud / Performance / Velocity 2011

Velocity didn't disappoint. Given my inner geek it's not surprising that this geeky discussion fascinates me, but I think the impact is much broader. The center of the discussion here is and always has been the largest Web sites, but I think the topics are important topics for many just shifted forward in time.  With a move to public and private clouds, everyone needs to be more sensitive to issues of performance (it becomes more challenging with remote data centers) and operating complex systems (understanding why an application performs poorly is much more complicated on a virtual fabric because resources are shared). These large Web sites have been living that reality for years. The principals in the discussion are very smart. Perhaps more importantly, the Velocity meeting has been fundamentally built aroud the notion of a community of practitioners and the value of sharing. So the discussion is largely unvarnished and the spokespeople are the experts (not marketers).

Not only are these Web systems remarkably complex and performance sensitive, the development methods are as agile as you can get. Both Amazon and Facebook deploy 1,000 or more new significant pieces of functionality every day without the benefit of long and complete test cycles. An essential value of the Cloud is the ability to create shared, multi-tenant applications and then drive evolution as fast as possible. This is vastly superior to long development cycles in large part because the only way to really learn about what your customers want is to give them your best shot and iterate and the value is as great in enterprise B-to-B applications (it's how IT helps the business move agilely and ahead of competitors).  Functionally, Facebook and Amazon evolve at warp speed. They represent the ideal that every enterprise IT organization strives for: rapidly evolving, largely automated computing. What makes it all the more interesting is that they do this at gargantuan scale, so it's not an issue of academic speculation that may not scale to enterprise levels.

If there are high-level truths to be abstracted from this year's conference, they might be data and dev/ops.  The key to making this all work is having lots of data, watching what's going on, improving performance and cost performance, and sensing and reversing updates that go bad. Dev/Ops, a unified organization, captures the wisdom that you can't do this with functionally divided development and operations. In the good old days enterprise software was developed and tested, and then thrown over the wall to be deployed and operated why the dev team attacked the next 18 month cycle. For truly agile and high-performance Cloud operations there can't be walls. There is still important functionalization but it all happens as one big team, especially under duress (had lunch with an old friend in the middle of Facebook and believe me, it's interesting but non-stop).

So I think the content of this meeting is highly informative to many more than the high-performance Web operations. For anyone contemplating making serious use of the Cloud this is a great peak into the challenges and the opportunities to some. More importantly it's a great way to see how IT organization has to evolve over time, with enough time to do something about it.

My hat continues to be off for the Velocity program team!