HP has just announced it is delivering a range of virtualization and cloud-computing capabilities it brands as “Next Generation DataCenter (NGDC).” The crux of the announcement is that capacity will be organized into a more abstract service layer, whether it comes from within the enterprise itself, or from HP’s latest on-demand offerings.

This is a vision, actually, that has been talked about for years by the large infrastructure vendors. I remember 10 years ago when Sun’s irrepressible Scott McNealy talked about delivering compute capacity via a “Big Freakin’ Webtone Switch.” Around the same time, IBM began talking about building server farms that will deliver capacity on demand to customers from remote locations.

HP’s “Adaptive Infrastructure as a Service” (AIaaS — now there’s a mouthful) is particularly interesting, since its a data center offered through the cloud from HP. HP AIaaS offers customers access to HP-owned and managed data centers that deliver applocations such as Exchange and SAP. HP did not mention others, though it did leave the door open to “other critical business applications.”

HP also is offering a range of virtualization capabilities which it calls HP Data Center Virtualization services, along with virtualization software, which supports all kinds of virtualization across servers, storage, networks and applications.

The move to cloud-based data centers continues, and thanks to high bandwidth networks, is technically feasible. Today’s economics is also helping to fuel this decade-old dream: the costs of building and maintaining data centers that can keep growing with demands for more data and digitalization of processes are going through the roof. Energy is one cost drain, but simply being able to find, train, and retain staff talent required to keep things going is a huge budget item. Tapping into data centers in the cloud offers a more incrementally priced alternative.

We just wrapped a survey of members of the Oracle Applications Users Group — which includes IT directors, CIOs, IT operations managers, and developers — on the extent to which they’re able to employ automated processes to manage multi-application workloads. In other words, are there processes in place in handle the management and movement of data across Oracle, PeopleSoft, and anything else that helps keep the enterprise running?

The survey of 344 companies, sponsored by UC4, found there’s still a lot of work to be done in this area. More than 75 percent of organizations surveyed are having difficulties managing and monitoring workload.

Over two thirds of survey respondents reported delays in business processing that could be directly attributed to the inefficiencies and errors that are introduced when manual intervention is required to drive processes that span multiple applications. Automation is helping address some IT process management issues, but adoption is slow. Three out of four operations still either rely on manual scripting, or simply don’t know what approaches they should take.

The need to automate IT processes is pressing. Not only are data center infrastructures becoming more complex; the challenge of managing workload that spans multiple business applications, such as Oracle EBS, PeopleSoft and SAP, adds to this complexity. The survey shows that organizations recognize the overall value and benefit that can be derived from end-to-end process management and visibility. Unfortunately, budget constraints combined with some companies’ inability to adopt standardized IT procedures is holding back more widespread deployment of IT automation solutions.

Most organizations turn to IT process automation to reduce the amount of staff time consumed by routine or manual tasks. However, in many cases, there is little room in corporate budgets for extensive business process management (BPM) solutions that provide automated modeling, management, and monitoring capabilities. One third of those surveyed are considering strategies based on Oracle Fusion, BPEL and Application Integration Architecture (AIA). However, at this stage most are undecided or unclear about the viability of these technologies as workload automation platforms.

Copies of the full survey are available here (registration required) at UC4’s Web site.

Just when you think the operating system space is about as saturated as it could get — and all the action has moved to the middleware and applications layers of the stack — there’s a new OS added to the mix.

Unlike most OSes, however, the new OS, called Cosmos, doesn’t run on the bare metal of machines. Rather, it’s a virtualized OS that runs within the .NET container, intended to support .NET-based languages (particularly C#) that are also resident on the .NET container. Cosmos stands for “C# Open Source Managed Operating System.”

The OS is also open source, issued under the BSD license.

Why develop Cosmos? Actually, Microsoft also has a .NET-based OS called Singularity that also runs within this environment. The reason stated in the Cosmos FAQ section is that:

“Primarily because it’s fun. But beyond that, how else can you boot .NET on a floppy or small USB stick? Who else will try to put .NET on the Wii, OLPC, and iPhone? We are also developing a TCP/IP stack. Imagine instead of deploying half a dozen virtualized OS’s, deploying many dozens of dedicated OS’s. One that only does DNS, a few that only do HTTP, etc. One instance, one function.”

Welcome to the party, Cosmos. Mary Jo Foley, who first uncovered this announcement, also spoke with Chad “Kudzu” Hower, a former Microsoft employee, who admitted the OS has a “super geeky” flavor to it, and even has garned independent interest from Microsoft staffers.