The lack of buzz about the recent VMware announcement of vCloud Express is, to me at least, puzzling. VCloud express is VMware's riposte to Amazon Web Services—a publicly available, get-started-with-a-credit-card, IaaS offering. Since VMware is a software company, it has partnered with service providers to deliver the product to market. In essence, vCloud Express packages up VMware technology, surrounds it with a set of APIs to control and manage the package, and allows service providers to offer vCloud Express—it's sort of cloud in a box.

I've been playing with vCloud Express and hope next week to offer some initial technical impressions, but in this post I'd like to focus on a cost comparison of AWS and vCloud Express (I think from this point on in the post I'll call it VCE to reduce typing).

[For timely cloud computing news and expert analysis, see CIO.com's Cloud Computing Drilldown section. ]

Overall, the pricing of VCE seems pretty comparable, and competitive, to AWS. This, to me, is a bit of a surprise. Most public cloud providers focus on using low-cost kit, often designed to their own specs. Google, famously, uses what appears to be a bare motherboard with components Velcro-ed on. VMware, at least in its internal data center manifestations, is typically provisioned with high quality (and high price) components from topflight suppliers like HP, EMC, and Cisco. When I heard rumors of the forthcoming VCE, I wondered if the use of componentry such as these would force VCE pricing above that of AWS. Thus far, the evidence is that VCE pricing is delivered to market roughly equivalent to AWS. And this pricing does not appear to be achieved by substituting cheap componentry in place of the expensive stuff common in internal installations. Terremark, one of the launch service providers (here in the US, the others are BlueLock and Hosting.com), notes in its specification pages that its VCE offering is based on Cisco and fibre channel SAN storage.

So, how does the pricing break down? Naturally, it's impossible to do a direct apples-to-apples comparison between VCE and AWS, but here is a rough comparison:

I have heard that, unlike AWS, VCE charges for the virtual machine whether it's running or not. On the other hand, when an AWS machine is not running, its image must be stored in S3, which incurs storage charges. Depending upon the size of the machine image, the total for a given machine might be equivalent for the two solutions.

All of the VCE service providers also offer private cloud computing services in a more traditional outsourced fashion; that is, cloud computing (again, based on VMware) hosted on the service provider's equipment. Unlike VCE, however, these other forms of cloud computing are not offered in an on-demand fashion, available for anyone with a credit card to begin using. These services are sold and supported much more like the traditional outsource relationship—heavy sales efforts, long-term customer commitments, focus on SLAs, support for critical compliance regimes like SAS70, and so on.

From my conversations, it seems that the VCE providers are hosting it on the same general infrastructure as their private cloud, although (perhaps) segregated onto separate VLANs and using different storage instances. My interpretation of this is that the VCE offering might pick up the compliance regime support that is in place for the private cloud offerings (i.e., the VCE would also be SAS70 compliant), though this remains to be confirmed. If this compliance support actually does transfer to the VCE offering, it would be a real user benefit, since many customers would prefer (or require) support for important laws and regulations before an application can be hosted somewhere.

Another benefit of VCE is that it provides (at least potentially, if only in the future) the ability to migrate applications and even entire cloud topologies between an internal cloud and a VCE-based public cloud. Of course, transferring virtual machines doesn't necessarily imply transferring the data associated with the VM, which poses its own range of problems. Nevertheless, a coordinated VM/storage transfer to a VCE cloud could be possible in the future. As you might know, I've been pretty dubious about the so-called "cloudburst" model, recognizing its limitations in terms of VM image formats; an all VMware infrastructure, both internal and public, might very wall aid this.

On the other hand, AWS has tremendous momentum, and VCE might not prove sufficient to displace it—particularly since there will be numerous VCE providers, which will spread the focus and make potential users have to consider more variables: not only what technology to use, but which provider. In my experience, providing too many alternatives can lead to customer paralysis, which might, paradoxically, turn people toward AWS. It will be interesting to see how it turns out, but from the perspective of a customer, having an alternative to AWS is good; competition generally is a positive for buyers.

Overall, VCE represents a vital response by VMware to Amazon; absent an on-demand cloud offering, VMware might find itself boxed out of an important segment of the cloud market despite its undoubted cloud in the internal data center cloud variant. As just noted, uptake of VCE will depend not only on its functionality, but how effectively VMware markets it, as well how well its service provider partners market it. Assuming those stars line up, I believe VCE can be a real contender in the on-demand cloud market segment.

Bernard Golden is CEO of consulting firm HyperStratus, which specializes in virtualization, cloud computing and related issues. He is also the author of "Virtualization for Dummies," the best-selling book on virtualization to date.

Follow Bernard Golden on Twitter @bernardgolden. Follow everything from CIO.com on Twitter @CIOonline.