In-Depth

What Pat Gelsinger Should Announce at VMworld 2016

Our experts put themselves in the CEO's shoes for the keynote address.

• By Keith Ward
• 07/21/2016

There's no doubt that VMworld 2016 will be one of the most important ever for VMware Inc. With the looming buyout of parent company EMC Corp. by Dell Inc., and challenges from every side, a clear vision for where the company heads from here will be crucial for allaying fears and cementing trust among the IT public.

That vision will be laid out chiefly by CEO Pat Gelsinger during the keynote address at VMworld. Given the crucial nature of this speech,Virtualization Review rounded up a panel of virtualization veterans from all corners of the industry, from vendors and journalists to administrators and analysts, to get their take on what Gelsinger should announce.

This isn't a prediction of what Gelsinger will announce; instead, it's advice on what directions VMware should go in to retain its place as the king of the virtual hill.

VMworld is one of the tech world's most prominent events, particularly for those immersed in IT infrastructure, annually drawing more than 20,000 attendees to what's evolved into a celebration of all manner of enterprise cloud technologies. With VMworld providing executives their biggest megaphone of the year, CEO Pat Gelsinger is likely already pondering how to address the following disquieting events since his last keynote:

• The imminent acquisition of VMware's effective parent and majority shareholder, EMC, by Dell; and resulting questions about VMware's future vis-a-vis the combined entity.
• Wall Street's extreme uncertainty about VMware's business prospects along with confusion around the VMW tracking stock, including the prospect raised by an SEC filing that Dell might buy up to $3 billion of VMware shares as part of the EMC deal. Together, the confusing news has led to wild fluctuations in VMware's stock price over the past year.
• Whether related to the Dell deal or not, the effect of departures by several key senior executives, including Gelsinger's right-hand man, longtime President and COO Carl Eschenbach; network virtualization guru and godfather of NSX, Martin Casado; head of vCloud Air, Bill Fathers; and CFO Jonathan Chadwick. All have left within the past few months, which has fueled speculation about Gelsinger's future at the company.
• The on-again, off-again plan to spin out Virtustream into an independent cloud provider targeting hybrid architectures by incorporating elements of vCloud Air; a prospect, since ditched, that could've bolstered vCloud's weak market position compared to other IaaS competitors.
• The rapid enterprise acceptance and financial growth of public cloud services, reflected in both AWS and Microsoft Azure growing revenue at high double-digit rates, with each running at a $10 billion annual rate and underscoring the legitimate threat to VMware's core business from Infrastructure as a Service (IaaS).
• Breathtaking technical progress in and adoption of application containerization software and associated clustering and orchestration systems, which could change how enterprise applications are designed and deployed.

These issues, together with the market saturation and concomitant slowing of VMware's core hypervisor business -- with compute license bookings declining 10 percent year-over-year -- mean Gelsinger faces the most challenging strategic environment of his tenure as CEO. Because VMworld is the company's premier event, when the spotlight shines on Gelsinger's keynote, customers, channel partners, technology and financial analysts, and (non-EMC) shareholders will expect answers.

At any vendor conference, the CEO keynote is somewhat formulaic: They celebrate the past year's success, portray a promising strategy for the future and express appreciation for the gathered customers' continued business. Within this template, Gelsinger's task is threefold:

1. Address competitive threats and product-line transitions.
2. Articulate opportunities and strategic programs based on core technologies like Photon/Photon Controller, NSX, VSAN, CloudFoundry and vCloud/vRealize.
3. Appease VMware defectors and doubters, such as those leaving for cheaper or built-in alternatives like OpenStack/KVM, Nutanix Acropolis and Hyper-V, or looking at various container technologies like Docker and Mesos for next-generation applications, migrating to Amazon Web Services (AWS) or considering Microsoft's compelling hybrid cloud alternative.

These are interrelated topics that Gelsinger can and should address by rebooting VMware's Software-Defined Datacenter (SDDC) strategy for the competitive realities in which cloud stacks are seen as a revolutionary departure from -- not an evolutionary extension of -- legacy virtualization infrastructure. Although the vRealize management suite can target external cloud services like AWS or OpenStack, these are still treated as extensions of a vSphere-centric world based on VMs and storage volumes, instead of a rich portfolio of higher-level application services.

VMware's cousins at EMC discussed an emerging dichotomy that exists in many organizations between infrastructure and management software designed for legacy applications from the client-server era, and those built on cloud principles using shared, on-demand, distributed services. VMware's roots are in providing virtual infrastructure for the former, but today's most innovative applications and online businesses are based on the latter. At VMworld, Gelsinger needs to elaborate on the messages from EMC World by describing in more detail how VMware intends to bridge the gap between the pre- and post-cloud application eras.

It's time to more clearly articulate the nascent strategy it revealed last year with VMware Integrated OpenStack, Photon and the Photon Controller, fill in details on how VMware will play in the world of open source cloud and container stacks, and clarify its product roadmap vis-a-vis the Native Hybrid Cloud product announced by EMC. Gelsinger should also outline the role of Cloud Foundry in this cloud native world: Will it remain a bolt-on, hypervisor-agnostic Platform as a Service (PaaS), or will VMware build tighter integrations between its infrastructure management software and the Cloud Foundry service fabric?

VMware's answer for the public site of hybrid cloud is the vCloud Air Network; and while that ecosystem has certainly grown, as evidenced by the IBM agreement in which Big Blue will offer vCloud Air, the actual usage of public vCloud services is unclear, but likely nowhere near the adoption of AWS or Azure.

Now that Amazon breaks out AWS revenues and Microsoft has become more granular in its cloud reporting, I'd like to see an accounting of vCloud Air revenue and usage, both from VMware services and that of its partners. With so-called hybrid cloud revenue, the majority of which is vCloud product plus Air services, only accounting for 7 percent of VMware revenue, Gelsinger needs to show how it can maintain current double-digit growth and compete with much larger public cloud competitors.

VMware's cloud strategy has focused on virtualizing and deploying to shared services legacy workloads designed for the pre-cloud era of client-server computing. At VMworld, Gelsinger must explain how VMware will get customers to think of it for Greenfield, cloud-native applications based on an entirely different technology stack and VMware's role in providing that stack.

Gelsinger must navigate a minefield of open business, market, product and technical questions at VMworld. Success means that VMware customers, analysts and shareholders have a clear idea of the company's future in the post-Dell world, an understanding of VMware's role in providing hybrid cloud technology without requiring a monolithic, vertically integrated suite of VMware products; how it can compete with a tightly integrated hybrid architecture from Microsoft in Azure and Azure Stack; and how the company will enable cloud-native applications.

The pressure is certainly on to deliver a rocking keynote, especially this year with the Dell/EMC acquisition looming large and creating a lot of uncertainty in the market.

1. First (and this one seems obvious), VMware must not turn its back in any way on vSphere. This is the bread and butter, with likely more than a 70 percent share of the enterprise virtualization marketplace. Like EMC's hands-off approach, Dell can capitalize on the brand association and marketplace leadership, but should keep well out of VMware's strategy and operations. The last thing VMware needs is Dell's factory/consumer model approach stifling innovation.

And regardless of any of the opportunities mentioned here, nothing will happen overnight. VMware needs to strongly commit to enhancing vSphere for existing customers and continue to broaden its appeal to attract new customers. This requires competing fiercely with Microsoft's Hyper-V and open source hypervisors that encroach on VMware's turf.

One example might be continuing to separately evolve VSAN for vSphere customers, even at the apparent risk of competing with EMC's ScaleIO (a current positioning challenge) or any new Dell/EMC software-defined storage solutions.

2. While VMware needs to protect home base, it will also need to embrace the inevitable. Hypervisor technology is commoditizing from basic functional levels upward. Both open source and third-party solutions have proliferated, and many offer significant alternative values. VMware needs to double down on enabling third parties to run very efficiently in its environment, and open up more to let them participate as first-class citizens. VMware will need to work hard with its broader ecosystem partners to produce more vSphere integrated solutions, like the recent SanDisk FlashSoft that leverages VAIO filter drivers to great effect.

Note that this means maintaining separate relationships with the vendor community. VMware can't afford to let the Dell/EMC deal impact its great technology partnerships with just about every vendor in IT today. It might even mean open sourcing parts of its core stack that could encourage competition with other Dell/EMC portfolio solutions.

3. VMware has been transforming year after year into an enterprise (OK, "cloud") solution vendor. Software-defined enterprise resources and cloud management, cornerstones of this transformation, need to be applicable across non-ESX environments.

For VMware, we think the current drivers for successful enterprise virtualization sales are NSX and VSAN. VMware needs to keep NSX evolving fast, and not let it get locked in to VSAN or vSphere anchor points. It would be great if VSAN could unlock from vSphere, but we think it's fundamentally married at this point. At least, however, VMware should develop NSX as a separable piece, like cloud management.

Cloud Management? Of course, vRealize is maturing into something much more than a virtual cluster manager. It now enables the virtual admin to broaden out and compete with platform/silo designed enterprise IT management solutions with rich features like intelligent key performance indicators, blueprints and integrated planning that fuel agility and even adjust predictive alerting, and native costing features that help tame and optimize IT spending. But vRealize is now also bigger than vSphere and vSphere clouds. Other VMware solutions can and should follow suit to provide cross-cloud agility and hybrid cloud migration and building.

4. Which brings me to my fourth point: It's going to be a hybrid cloud world. We think Airwatch and other client offerings should be floatable across many different underlying cloud platforms. Somehow, vRealize Orchestration and distributed resource scheduler/vMotion should come together as a service that could fundamentally extend to help on-premises ESX workloads seamlessly hybridize to public clouds.

In the longer term, VMware should strengthen all offerings above the hypervisor for deployment in public clouds, maybe with some kind of portability layer so that even vSphere-based workloads can run across those other clouds with ease. Note that VMware supported containers early on, and that experience might offer some clues to how to best support an increasingly agile, distributed, hybridized enterprise.

5. Finally, VMware needs to stop fumbling with hyper-convergence and focus on software, not hardware. Let Dell/EMC run with converged infrastructure solutions (and just maybe they'll find their own actual hyper-converged solution soon). EVO was interesting, but ultimately it was just an exclusive hardware compatibility list (HCL) that limited partners, where instead it should have really been an open recipe book that encouraged solutions. In fact, the point of virtualization and software-defined resources is that the hardware shouldn't actually matter (functionally, although there will, of course, be large differences in performance, reliability and TCO).

Gelsinger should offer some advice to Dell along the lines of, "You are now lagging behind HP's Synergy ‘bare metal cloud enabling' infrastructure offering, and need to step up soon." Along with the onrush of containers, this new dynamic physical infrastructure is another reason it might be a losing strategy to peg the future on proprietary hypervisor-level functionality. VMware should already be planning how to extend management across hybrid virtual and dynamic physical ("composable") infrastructures. For example, by leveraging HPE Synergy APIs it should be a simple matter for vRealize users to now orchestrate and deploy blueprints, either virtually or physically on demand.

We see VMware and RedHat Inc. coming to blows in the next few years. RedHat has been coming at this from the "other side," starting with open source initiatives and building enterprise values on top with great success quarter after quarter (for many years).

VMware has been coming at this from the successful commercial hypervisor; after 15 years of ESX, it dominates the enterprise market. But it's now quickly becoming a commodity marketplace. In the next five years, the enterprise IT spend will shift toward higher end software-defined resources and pre-converged (hyper-converged, et al.) solutions. VMware has all the right elements that can be martialed, but will need to make some bigger pivots away from a monolithic platform vision to remain on top of the game.

Finally, VMware is sure to be looking at megatrends like IoT (fast streaming, edge-produced data), ever-bigger data (data lakes and so on), always-on data processing (24-hour, follow-the-sun distributed processing; no time for backups), completely hybridized infrastructure and dynamically agile management requirements, and massive scale-out clustering.

Like much of the world -- especially the non-American world -- I live in a boom-and-bust economy. My clients tend to be very OpEx-sensitive: Buying during the boom and sweating assets through the bust is the norm.

I know how tight times can get for the majority of the world's businesses, and I know that subscription fees can kill a business. IT, from infrastructure to applications, can only be scaled back so far. One bad season can mean the end of a business if it can't sweat assets.

Of course, this doesn't apply to every business. A lifetime in the trenches has made me an advocate of "the little guy"; I disdain vendors that focus only on the Fortune 2000, but am aware of the disproportionate revenue generated by large organizations. Smaller organizations also cost more in terms of support per license sold than larger ones.

When talking about the future of VMware, all of what I've just listed matters. In my opinion, it matters rather a lot.

In my unicorns-and-rainbows-filled happy world I really have only three requests of VMware. These three requests are focused on solving the needs of myself, my colleagues and my clients as they exist today. The future is the future, but my wish list solves the now.

First up, I want a VMware version of what Microsoft's TechNet used to be: Pay a small fee and get access to a bucketful of free software. Enough software to run your whole lab, be that at home or at work.

There are a few reasons I believe this is important. The first is that VMware's success rests in large part on the community that has grown around it. Most of that community simply can't afford to buy licenses for VSAN, NSX, vCD, vRealize and so forth without a solid business need. It can't prove that business need unless and until it understands the software, and for many people that understanding comes from doing.

The sourer note driving this request is simply that VMware has very real competition today. Microsoft is no slouch, and open source alternatives ranging from the hard-to-use OpenStack to Scale Computing, Yottabyte and NodeWeaver all offer compelling virtualization stacks at rock-bottom prices.

The classic argument against this is that once a TechNet-style licensing option is available, small businesses will simply buy that and "pirate" the licenses by using them in production, instead of restricting usage to the lab.

I question if this really matters. VMware has repeatedly proven it doesn't care much for small customers anyway, so it can't be bringing in too much revenue. Better to keep these companies in the VMware fold so that the few that do grow eventually buy lots of VMware licenses, instead of getting comfy with the competition.

This leads me into the second and third requests: I'd really like to rejigger the Essentials Plus package. It should be enough software to run the full suite of VMware software (NSX, VSAN, cloudy things and so on) on three servers for cheap enough that your average 50-man shop won't have trouble affording it.

I'd also like to see a smoother transition from Essentials Plus to traditional licensing that doesn't functionally entail throwing away the investment made in the Essentials Plus licenses... and preferably isn't such a huge jump in costs.

The rationale behind this is the same as earlier; with viable competition in existence, it's better to keep the milled masses in the fold than see them explore new options.

While I want VMware to focus on solving the real-world problems of the majority as they exist today, VMware desperately needs to focus on the seemingly overwhelming problems of tomorrow if it wants to survive. VMware is facing competition on multiple fronts, and it needs an unassailable niche from which it can launch forays into adjacent markets.

Virtualization is emphatically not VMware's unassailable niche. Depending on how you slice the data, VMware might still be the dominant provider of virtual infrastructure, but that won't last long.

Microsoft has a critical trifecta: public cloud, private cloud and the bits that marry the two. Amazon, IBM, Google and others have great public cloud offerings. OpenStack is no longer a science experiment, with companies like ZeroStack offering on-premises hyper-converged appliances, services providers around the world lighting up public offerings and HPE willing to build you anything you want.

If that weren't enough, basic virtualization is a niche that can be filled by open source-based providers for a fraction of VMware's costs. VMware may be able to win on price against Nutanix on the high end, but it gets wrecked by Scale Computing and others on the low end.

This leads me to the niche that VMware needs to carve out for itself: ease of use. One of the biggest reasons VMware rose to the position it is in today is that it was significantly easier to use than the alternatives. Combined with excellent community management, this created an army of fanatical zealots that evangelized VMware around the world.

Today, VMware's competitors are easier to use. Some are easier to install and configure, some are easier to administer and some have better end-user-facing self-service portals and marketplaces. Some competitors even have better versions of all three.

We've long since passed the point where the features needed by the overwhelming majority of customers exist. Now it's time to go back to square one and rethink how those features are engaged with, how automation works and how to ingrate the whole of VMware's intellectual property into a single, simple offering.

EMC and NetApp once got cocky about ease-of-use, too. They were comfortable requiring specialists to manage their products, and those specialists liked that status quo. Then the storage industry exploded with innovative startups. Margins shrank, market share was readjusted and the once-complacent behemoths are now forced to adapt or perish.

VMware should cede the public cloud niche to Amazon, Google, Microsoft and IBM. It isn't going to do better in that space. But on-premises? Creating the infrastructures upon which regional services providers base their public offerings? VMware can -- and should -- win here.

If it wants to survive, VMworld 2016 should see VMware announce a company-wide simplification process. One UI for all products. Simplified licensing. Above all, it needs to harmonize the release cycles of its software offerings ahead of a true integration into a single, fused product.

There is no way VMware can deliver on that massive a goal by VMworld 2016, or even VMworld 2018. But knowing the project is underway matters. It may well prevent a lot of companies from deciding that the cost of managing and maintaining VMware is too high and exiting the VMware ecosystem.

Let's see what news late-August brings.

While I can't predict the future, I think the focus should be on what I see as the five growth opportunities in the market today: end-user computing, software-defined storage and hyper-convergence, public cloud, security, and analytics.

In today's fast-paced, always-on society, the importance of the end user experience cannot be underestimated. VMware should double down on its investments from the past year and focus on making the virtualization platform for desktops and applications more integrated, smarter, faster and cheaper. How? By remembering that the competition is not Citrix Systems Inc., it's the physical PC. Every modern PC has flash (SSD) and graphics (GPU), so why settle for less in the virtual workspace?

In a market of 1.5 billion enterprise desktops, fewer than 10 percent of PCs are virtualized. We can, therefore, expect a huge growth market focused on the virtual workspace, which I define as virtual desktop infrastructure (VDI), virtual workstations and remote applications.

VMware should support and encourage the virtual workspace ecosystem. For example, NVIDIA has worked hard to drive development opportunities for shared graphics in servers to increase performance for the virtual workspace. Similarly, VMware should team up with Intel Corp. and AMD to bring more GPU options to the virtual workspace and partner with software-defined storage (SDS) vendors to create more efficient, high-performance and flexible storage solutions. Over the next few years, the desktop virtualization market will become more competitive, producing a wider variety of customer-facing solutions with lower costs and greater flexibility.

The key to increasing adoption of virtual workspaces is to look for more opportunities to deploy SDS andsof ware-defined networking (SDN), and in the process bring more agility, security and performance into the data center and cloud infrastructure. CaPex and OpEx costs will be driven down by harnessing the power of processors, memory, flash, commodity hardware and intelligent software to create a more automated, agile and programmable infrastructure that can quickly adapt to changing market conditions.

A key technology for all of this software-defined infrastructure is REST APIs, which will enable IT to manage multiple virtual machines (VMs), containers and cloud instances from an integrated management solution. REST-based APIs will be a must-have technology; if a solution doesn't have a fully open and documented REST API, the solution isn't suitable for the modern workspace and datacenter.

VMware has made some inroads with vCloud Air as a hybrid cloud platform that enables customers to move workloads to and create applications in the public cloud. It is built on VMware vSphere, so it would be familiar to cloud architects and developers alike. Apps that run on vSphere on-premises will also run on vCloud Air with consistent performance and functionality.

To increase its footprint in hybrid and public cloud, VMware could look to a strategic alliance, such as with Google, to fill the void. This combination would promote competition with Azure and Amazon.

Every day, the media reports on companies that have had files and information stolen or compromised. The business consumer is in constant danger of being targeted by hackers and malware. Ransomware can impact everything including data, a brand or even an identity.

To address these growing concerns, VMware should look at making smart investments to deliver capabilities that are lacking today. It should integrate smart endpoint protection and security solutions, such as Bromium, machine learning and analytics in the cloud, identity management and user-friendly collaboration solutions to automate identification and resolution of security threats before the damage is done.

VMware can also enhance the UX by applying research and development to extend monitoring solutions that increase visibility to the remote endpoint, gateway and Software-as-a-Service (SaaS) applications. VMware can be the virtual desktop or the remote application.

VMware's vRealize Operations Manager provides great insights in datacenter operations, but is lacking in monitoring for the end-user experience; the company also needs to deliver analytics and diagnostics that can be relevant for the virtual workspace.

We live in two worlds: the slow, static PC world and the agile cloud world. VMware needs to bridge the two; from desktop computing and software-defined datacenters to public cloud and cybersecurity, it all comes down to the end-user experience. Industry leaders like VMware can add business value and accelerate market opportunities by understanding the business consumer's evolving requirements. We can all drive change and prioritize the end-user experience. We just need to step out of our comfort zone and do it.

1. A Less Intimidating Product Lineup One of the most important things that any business can do is to attract new customers. However, VMware's inherent complexity might be interfering with its ability to attract new customers. Once upon a time, VMware's only offering was its hypervisor. Over time, of course, VMware has developed numerous products and services. There are so many, in fact, that those who don't have a lot of experience with VMware products almost don't know where to begin.

One of the first things I'd do as VMware CEO would be to make my product line much less intimidating to prospective customers. Creating a simple infographic illustrating what the various products do, and how they're related to one another would be a good start (assuming that VMware made the document easy to find).

Another thing my company could do to help new customers is create a wizard-like interface that could assist them with the planning and purchasing process. Such a wizard could ask customers various questions (written in plain English, without unnecessary technical jargon), then provide an on-the-spot quote to the customer. The quote would tell the customer exactly what licenses they need, what functionality is provided by each license, and how much it will all cost. Acronis provides a similar wizard to help its customers, and it works very well.

2. Strategic Partnerships for More Cloud-Like Application Deployments
As CEO, I'd work to build strategic partnerships with major application vendors, with the goal of simplifying the application deployment process. VMware has already simplified the management of SaaS applications, but it needs to take the next step and focus additional attention on applications that will reside on-premises. While this might seem counter-intuitive, given the world's "cloud everything" mentality, the long-term cost of SaaS has some organizations reconsidering on-premises applications.

Public cloud providers such as Amazon and Microsoft have long provided customers with self-service portals that allow fully configured VMs to be automatically deployed. Although such functionality is already possible in VMware environments, imagine how much easier an administrator's life might be if VMware provided pre-built, optimally configured, multi-tier application server templates as a result of partnerships with application vendors.

3. Software-Defined Data
VMware's software-defined datacenter (SDDC) was revolutionary, but the one thing that VMware has yet to virtualize in a meaningful way is the data within the datacenter. VMware has long stressed the importance of business agility, defined as a business's ability to rapidly deploy new services on demand. VMware is in a unique position to greatly enhance business agility through data virtualization.

Consider an organization that's 100 percent virtualized on VMware. Such an organization would likely have a wide variety of applications running within VMware VMs. The application data might even exist within VMware datastores. With this in mind, imagine the possibilities if VMware were to implement a level of application awareness that allowed a dedicated tool to not only read the data, but also to understand what types of data exist within the organization. In other words, application data is virtualized and treated as a pooled resource.

This pooling of application data probably wouldn't benefit any of the organization's commercial applications, but it would open the door to deep data analytics and business intelligence. It would also make it possible for VMware to create a tool that would allow an organization's in-house development staff to easily build custom applications that leverage data from the organization's existing applications. It might even be possible for such a tool to include a drag-and-drop interface that allows developers to combine data fields in any manner that might be required. Because the software would have deep insight into the organization's applications and the corresponding data, it should be theoretically possible for such a tool to honor existing permissions.

VMware needs to consider the competitive threats to its vast product lineup, including:

• Citrix: Offers products in all seven layers of the Kusnetzky Group model of virtualization technology (bit.ly/1tuF2ws). The company typically plays the "vendor neutral" and "Do you want so many eggs in one vendor basket?" cards in its marketing messages. Its access virtualization technology often supports new mobile device technology on the day it's announced.
• Microsoft: Microsoft has developed or acquired technology in all seven layers of the Kusnetzky Group model. Microsoft is increasingly baking virtualization technology into its OSes, development tools, management tools, storage tools, database and anything else it thinks about. For the most part, virtualization is being used as a tool to bolster the interest in and adoption of Microsoft products, rather than truly being focused on merely solving enterprise computing problems.
• Open Source: Open source communities also have addressed all seven layers of the Kusnetzky Group model. The technology coming out of these efforts often are well tested and quite sound, but require some strong skillsets and flexibility. The suppliers in these communities typically ask, "Why pay the x tax on your business?" (where x is whatever commercial supplier they're competing with at the moment) when speaking about any commercial vendor. Red Hat, for example, was known to speak about the "Red Hat discount" that other vendors would offer when they learned that Red Hat was a competitor.

There's also a lot of competitive pressure in cloud computing:

• AWS: AWS is the market leader and offers a worldwide presence. The company has grabbed the leadership position. It's driving forward with a strategy of proprietary technology and complex pricing, and driving down costs.
• Azure: Microsoft is a latecomer to the cloud game, but it's using the same tactics it has used for years to get into enterprise datacenters. That is, it builds tools into all of its products that drives decision makers to believe that if they're using Windows, the best choice for any other technology is Microsoft. The company's cloud offerings came out of nowhere and are now the second choice for suppliers.
• Open Source: Open source communities are supporting a number of cloud platforms and a number of variations on each cloud platform. As with open source virtualization technology, the technology coming out of these efforts often are well tested and solid, but require high levels of knowledge and skill. They will also point out that their tools are modular, modifiable and can be tuned to the precise needs of the enterprise.

VMware has been lifting pages from the Microsoft playbook. It started with VM software and has been extended out to all other areas of virtualization. Then the company started moving that technology into cloud computing environments. It has also selected Rails to be its enterprise workload delivery platform, and acquired technology allowing it to project into that market place, as well. Typically it uses a Microsoft-like "embrace, extend and control" set of tactics.

Once a company selects one of its products, they find that it's simply easier, and less of an integration headache, to select another one of the company's products in adjacent markets.

But much of the market is now demanding open, interoperable, extendable, reliable and cost-effective software. This often extends to demanding simple terms and conditions and easily understandable, believable licensing. They expect not only a low entry cost, but also expect low incremental costs as new features and functions are added to the portfolio.

Dell's acquisition of EMC and VMware muddies the waters quite a bit. It's not yet known how Dell is going to integrate enterprise storage and virtualization/cloud computing infrastructure into its portfolio. So, I'm going to ignore the Round Rock elephant in the room for a moment.

Here are a few suggestions for Gelsinger:

• Balancing a huge share of the market and a diminishing level of growth for the market leader often leads to internal demands for finding ways to increase revenue derived from today's customer base. This wouldn't be a good move in today's market; increasing prices in the face of extensive competition coming from other suppliers would cause many cost-conscious organizations to consider jumping ship.
• Building a walled garden in customer datacenters, while an attractive thing for product planners, flies in the face of customer demands for openness, interoperability, security and management.
• Make it easy for your customers to manage and secure multi-vendor environments that include some of your tools and technology. You know that they're going to do it anyway. If you make it difficult enough, you might find yourself written out of their future plans.