According to Neil Cowit, New York-based HPC solution specialist for Microsoft Corporation, Microsoft Windows Compute Cluster Server 2003 provides a secure, cost-effective solution for the compute-intensive requirements of financial services.

In a recent interview with Wall Street & Technology, Microsoft CEO Steve Ballmer said, “The way in which performance gains are achieved on computer chips has changed, creating a new challenge for HPC software. For years, Intel just kept doubling, doubling, doubling clock speeds. Now they don’t double clock speeds anymore – they give us twice as many cores or processors. Figuring out how to take any application and parallelize it so it can run across multiple cores will be a key, not just in high-performance computing and not just in financial scenarios but in all applications figuring out how to exploit the increase in power that physics is giving us.”

We checked in with Cowit for an update.

WFS: How has Microsoft’s HPC offering been accepted since its launch last August?

NC: Quite nicely, thank you. Then again, running compute-intensive work in a Microsoft-based environment isn’t new in financial services. A large Microsoft user base has existed for several years with some of the largest banks using Microsoft for their dedicated HPC-computing requirements.

What is new is Microsoft’s focus in this area in terms of personnel, product and partners. Third-party firms DataSynapse, Digipede and Platform Computing, all Microsoft partners, have large groups of customers. More recently, HP, Dell and IBM announced specialized sales teams and configurations that best fit their products, plus a number of third-party software firms have begun porting their products to Windows Compute Cluster Server.

WFS: What is Microsoft’s value to customers building their HPC environment with Microsoft?

NC: Running an HPC environment on Windows isn’t simply about technology. It’s about the ROI through what I call the economies of familiarity.

WFS: Meaning?

NC: Customers can leverage the skill sets of their established Microsoft operations and development teams to build, run and manage the applications of an HPC environment. Employees are able to access, manage and report on HPC results via a familiar, Windows-based environment, making them more productive by performing jobs and reporting the results via Microsoft Office programs.

We’ve also simplified deployment and management of an HPC infrastructure so you can come to one place and get an integrated environment of tools for HPC. It’s all about high performance and high productivity. For example, customers familiar with using Microsoft Operations Manager and related management packs to monitor the health of their IT systems can use the same tools for Windows CCS.

We’ve added familiar management tools, such as a parallel shell, to the Compute Cluster Pack to make management of Microsoft-based HPC environments even easier. Plus, developers can leverage the familiar environment of Microsoft Visual Studio 2005, building and deploying parallel, Microsoft .NET-based apps to meet changing business needs.

WFS: Does all of this apply only to capital markets?

NC: No, insurance and banking embrace HPC to refine their models to reduce risk, shorten time-to-market when creating new products and improve overall operations performance.

WFS: What types of workloads do you see people running on Windows-based HPC environments?

NC: We see attention focused on real-time derivatives pricing, risk management modeling and the like. For example, a European customer’s structured derivatives business built its second-generation HPC cluster using Windows CCS to speed pricing and risk management applications and simplify development and management – improvements that allow customers to increase the scale of their business and shift the focus of IT resources on solutions for that business.

WFS: Any other examples?

NC: We also partner with software and hardware vendors to get latency down as far as it can go. Let’s not forget all the behind-the-scenes batch work completed before the next business day, as well as those firms using HPC to migrate work off mainframes.

WFS: What are some of the HPC challenges customers face today?

NC: Well, parallel programming continues to be somewhat of a black art in financial services. The amount of heavy-lifting necessary to integrate an existing application to HPC can be considerable, but the rewards are high.

Microsoft added tools such as a parallel debugger into Visual Studio 2005 to speed the development process, proof that the industry must continue to develop software development tools to ease the task of parallel programming and Microsoft is making significant strides in this area. Firms like Digipede also have some rather exciting .NET development and deployment tools that speed integration.



WFS: Doesn’t Microsoft Compute Cluster Pack compete with existing vendors, including DataSynapse, Platform Computing and Digipede?

NC: No, they’re cream-of-the-crop in schedulers, Microsoft supports them and they run very well on Windows Server 2003 Compute Cluster Edition today. Interoperability is something that has been part of the design goal from the beginning. Looking back to Supercomputing 2005, Bill Gates spoke and demonstrated the sharing of work between different scheduler vendors and different operating systems in different parts of the country. At Supercomputing 2006, we demonstrated job scheduler interoperability with Platform Computing using a new industry specification.

And yes, interoperability also means sharing work between Linux and Microsoft HPC environments.

WFS: Is there a business model for systems integrators to be successful with HPC?

NC: HPC benefits every business in the Microsoft value chain. Our model’s always been built around the system integrator and application developer who bring a specialized capability to the table. Whether it’s departmental- or application-based deployment or a firm-wide utility-based deployment, applications do not magically integrate into HPC environments. Skilled developers are necessary, which is why we’re working with integrators Infusion Development, Vast Computing, Avanade and Polaris and other certified partners.

WFS: What’s most difficult in deploying Microsoft grids?

NC: The difficulties are not in deploying Windows. It’s in how you deploy and manage an HPC infrastructure within your company. You must understand the political nuances, some of which revolve around deploying a utility versus a silo approach, and in defining a charge-back mechanism that doesn’t run counter to your goal of cost-effective computing.

Customers also need to strike a balance between tight controls to achieve desired SLAs and providing flexibility for innovation.

WFS: Where do Excel 2007, Excel Services and HPC come into play?

NC: Excel’s the most well known, most utilized application in financial services. Thanks to the advancements in Excel 2007, multi-threading enables you to put Excel on steroids. The limits on a spreadsheets’ size has been greatly expanded while multi-threading allows you to set the number of threads you need to perform calculations in parallel.

Excel 2007 also identifies the parallelizable code areas the first time it runs through the sheet. The second time the sheet is run, those areas that can run in parallel will then run in parallel.

WFS: Take that a step further.

NC: User-defined functions can be created. Then these computations can be distributed outside of the desktop to Windows CCS.

Excel Services opens up a whole other world of opportunity. Being able to protect your IP, deliver a single version of the truth through a solution built on Excel using HTML and deliver performance by distributing computationally intensive work through user-defined functions to Windows CCS is a solution satisfying a lot of customers. If you work with parallel computing, incredible performance is within reach, and in many cases that speed can translate into profitability.

WFS: That’s a lot to absorb.

NC: There’s a lot to share….