Recent Microsoft (MSFT) and Gartner announcements have put hybrid cloud front and center for the next 3-5 years, and for perhaps a much, much longer period. Market caps for several players could be at stake: during these 3-5 years you could see the hybrid cloud eclipse the public cloud, both in terms of growth and overall enterprise spend. You think I jest? Read on.
I talked about the multi-billion market potential of the hybrid cloud in March 2013's Will VMware or Microsoft Cash in on the Hybrid Cloud (VMW). It could be a $60B total addressable market, when one makes a few reasonable assumptions:
A leading analyst firm forecasts the shipment of about 10M servers per year over the next three years, with about 50% of them being x86. What if enterprises wanted the strategic scalability, agility and protection of the hybrid cloud model, instead of the tactical agility, efficiency and protection of server virtualization?
Hybrid cloud would give enterprises more complete control of their apps and services.
If VMware could get $2k/year for each server (traditional and x86), that would amount to an additional TAM of $60B based on a three year refresh rate. Yet that would represent a major business model shift and limit the amount of lock-in that VMware would have over its customers operating on its private cloud platform. It could face margin erosion for its core lines.
These estimates only appear shocking because public cloud has been the darling of the press and Amazon (AMZN) has done an amazing job promoting the cloud, especially for new apps and as an agile Dev/Test environment. Yet recent steps by Microsoft, VMware and a host of OpenStack cloud players are likely to make 2014 a tipping point for hybrid cloud momentum. And with powerful new competitors (all battling over hybrid cloud leadership), Amazon risks getting sidelined from the buzz it has so adeptly built over the last three-plus years.
For more recent hybrid cloud background see Rachel King's Hybrid Cloud Refresh and Charles Babcock's coverage of Gartner's enterprise hybrid cloud adoption predictions. See also Jon Collins' perspectives at "Hybrid Cloud is dead; Long Live Hybrid." These have all appeared in recent weeks in response to Gartner and Microsoft hybrid cloud publicity. Yes, the hybrid cloud battle begins.
The real driver to hybrid cloud won't be price but agility according to almost a dozen key analysts I've spoken with since this summer's VMworld. That is consistent with enterprise cloud architects currently dabbling in various cloud services and platforms. Enterprises don't want to be stuck in anyone's cloud, however inexpensive, robust and powerful it is.
That suggests that the middle ground between the cloud and the data center may be where the hybrid cloud inflection point really is, versus the comparative attributes of say a dozen different walled gardens. Note my comments in Hybrid Cloud will be won in the Middle Ground (from September 2013).
The hybrid cloud pressures are building:
1) The case for data center construction is rapidly eroding for growing ranks of companies, except for a small handful of cloud-enabled power players with the in-house expertise and hundreds of millions in capital at their ready disposal;
2) Data centers are also becoming more expensive to manage as complexity increases with the evolution of IT and the powerful benefits of virtualization run their full course within legacy environments; and
3) The handful of power players mentioned in #1 are investing more than $50B per year in some of the most advanced data centers that have ever been built, often utilizing methods and clout (advanced designs, energy efficiency, tax advantages, etc.) that most companies cannot comparatively exploit.
4) As Amazon continues to promote the public cloud as the enterprise endgame, new, credible cloud service provider competitors will emerge with "close enough APIs, services and options" and deep pocket investment capabilities while giving enterprises hybrid cloud control over their apps and infrastructure. In other words, hybrid cloud automation enabled by integration.
Here is a chart prepared for CloudVelocity showing the hybrid cloud model adopted by one of its customers. Notice the authentication service integration between the data center and the cloud; the public cloud and the data center are treated as a seamless environment. I think this kind of deployment is the future of the hybrid cloud.
That makes the hybrid cloud migration and integration process, strategic to the evolution of hybrid cloud operating models. When those processes are fully automated we will see a hybrid cloud adoption hockey stick, a rapid rise in enterprise cloud adoption far exceeding today's hybrid cloud growth rate. Cloud migration needs to be fast and inexpensive, in order to support enterprise agility requirements.
In September I interviewed CloudVelocity customer QL2 Software, a Seattle-based data and analytics firm focused on a handful of vertical markets including travel and energy. They were able to reduce their cloud migration costs for a 50+ server infrastructure by as much as 75%. As cloud migration and integration solutions continue to mature (and support production environments) hybrid clouds could become the norm. In short, the middle ground of tools, services and experts required today for hybrid cloud deployment could be automated, with major impacts on adoption.
That then sets the stage for the emergence of Microsoft and VMware in the hybrid cloud wheelhouse.
Windows Azure: Crossing the Hybrid Cloud Chasm
Last month I spoke with Microsoft Corporate Vice President Brad Anderson, as he shared Microsoft's vision of a hybrid cloud platform, including recent announcements of Windows Azure Pack shipping in Windows Server 2012 R2. The Azure code will enable high scale hosting and management of web and virtual machines.
Within minutes it became obvious that Microsoft understood the critical importance of hybrid cloud integration versus simply offering a competitive IaaS to AWS.
Microsoft also has significant service provider expertise and massive footprint for Azure development as it distributes Azure into data centers on Windows servers. Microsoft has what Amazon is hoping to build out: access to most if not all of the world's data centers. Microsoft could deliver a powerful blow to Amazon in the enterprise space by having a more advanced, more enterprise-friendly approach to hybrid cloud integration. In short, they make it easier to build hybrid clouds.
If Microsoft can expand the Azure footprint faster than Amazon can convince enterprises into an "all or nothing" leap into their impressive clouds, Microsoft could easily be in a commending position. So could VMware, thanks to their core expertise and footprint in virtualized data centers and "private clouds," whatever you think that means. So in 2014 the cloud momentum could shift from the perhaps tired "public cloud purity" momentum driven forcefully by Amazon to a pitched battle between two very powerful titans who have been in the ring before, when Microsoft launched Hyper-V against VMware's hypervisor beachhead.
That is why I think that in 2014 we are likely to see a direct battle: x86 servers with VMware versus Windows servers preloaded with Azure. Note that Azure has been gradually gaining share since 2011 due to very aggressive pricing. Amazon risks watching this battle from the sidelines; especially if they continue to advocate the "all or nothing" public cloud.
Additional disclosure: CloudVelocity is a Microsoft Azure and Amazon AWS partner. Today it supports hybrid cloud automation for Windows and Linux apps running in AWS and has advised analysts of its intent to support Azure hybrid clouds in coming months.