Visit additional Tabor Communication Publications
July 01, 2010
Grid computing vendor Digipede is something of an enigma in the HPC world. The company has built its distributed computing offering, Digipede Network, atop the highly popular Windows/.NET platform, and in the process has become an unrepentant Microsoft booster. For traditional HPC users, anything not Linux plus MPI or OpenMP is mostly looked upon as an eccentricity, and in some corners, partnering up with Microsoft is seen as treason.
Even though Digipede hasn't embraced the HPC community's open source cultural value, the company has made it work. Founded in 2003, it remains one of the few pure-play grid computing vendors left. Because Digipede Network is built for .NET apps, and uses a different work management model than traditional job schedulers, it has little competition from other grid computing products.
The Digipede Network solution is rather straightforward. It consists of a Digipede Server and one or more Digipede Agents. The Server parcels out client requests to the Agents, which are responsible for managing the resources at the node level (servers or desktop clients). Like any grid computing solution, the idea is to distribute a compute-intensive job across multiple cores, processors and compute boxes, with minimal changes to the application source code.
The secret sauce for the company is the Digipede Framework SDK. It enables Windows .NET developers to take single-threaded, object-oriented apps and transform them into distributed computing programs. According to Digipede this can be accomplished with just a smattering of source changes -- about 20 lines of new code. That allows conventionally-trained programmers and their legacy applications to make the jump to distributed computing with a minimal amount of pain.
That may seem like the Holy Grail of programming, but Digipede picks its apps carefully. Rather than attempting to parallelize any kind of software, the technology is focused on what Digipede CEO John Powers refers to as "delightfully parallel" codes, or in conventional parlance, embarrassingly parallel. These are applications that can be easily split up into many tasks that have little dependency upon each other. This model of parallelization can lead to near linear speedups in performance. "That's nothing to be embarrassed about," laughs Powers.
Digipede Network runs on all modern Windows client and server platforms, from XP to Windows HPC Server, and is well integrated into the componentry of Microsoft's software empire. Besides the .NET framework itself (and Parallel Extensions), Digipede is interoperable with Visual Studio and all the server-side technology. Keeping up with the latest and greatest from Microsoft is a full-time job, but keeping the .NET developers happy is the prime directive at Digipede.
The company's latest release, Digipede Network 2.4, moves the technology up to .NET 4 and adds Windows 7 certification. The new version also includes API improvements, finer control of multicore computing, as well as performance improvements under the hood.
Although the company doesn't focus strictly on HPC-style technical computing applications, a lot of its customers' codes fall into that category. Probably the biggest single industry for Digipede is financial services, especially hedge funds. Companies in this vertical do market risk simulations for a living and the vast majority are on Windows platforms. Other big customers include defense firms and electric utility companies.
Up until 2007 especially, financial services drove the success of the company. When the industry imploded with the Great Recession, some of Digipede's most important accounts disappeared. "Our biggest customer in 2007 was Countrywide and they didn't make it to 2008," notes Powers. The hedge fund companies held their own through the economic downturn, though, and today the financial space still accounts for about half of Digipede's revenues.
The electric utility space is an emerging market for the Digipede, and Powers says this industry has taken off in 2010. The applications themselves -- risk management and market simulation models -- are much like those in the financial arena, and thus well-suited to Digipede-style distributed computing. System size is still relatively small though, generally in the 10 to 20 node range.
As a Microsoft Gold Partner, Digipede receives plenty of support from the software giant to help build its accounts, but its the Windows HPC Server offering that is aligned particularly well with Digipede's strategy. The HPC product is offered at a steep discount (compared to the standard Window Server) in order to be competitive with the corresponding Linux offerings from Red Hat and Novell. The only license restriction is that you can't run other Microsoft server components like SQL, Exchange, and SharePoint under the HPC banner. It essentially levels the playing field with Linux, cost-wise, in the high performance computing arena. "The HPC Server is the best thing that's ever happened to us because it's a way to buy Microsoft's best OS technology at 80 percent off," explains Powers.
Besides the Microsoft fixation, the other distinguishing characteristic of Digipede's strategy is its resistance to jump into the cloud computing space. Unlike many grid vendors that are expanding into cloud management, Digipede still views itself as a distributed computing vendor. According to Powers, there's very little going on in the cloud space to make it easier to deploy .NET applications right now. Even Microsoft has been very slow to talk about it, he says.
Azure may eventually turn out to be the avenue for .NET in the cloud, but Microsoft seems more interested in capturing new applications. "If I were building the next Facebook, I'd head for Azure in a shot," says Powers. "But if I had an existing legacy application that burned through a lot of compute cycles, I think Azure has a ways to go before it tackles that."
Digipede Network can run in cloud environments as well as it can run on dedicated clusters and workstations, according to Powers. So at least for the time being he seems content to dance with the one that brung them. "We view cloud as being a deployment choice as much as anything," explains Powers. "People talk about it as this profound revolution. In terms of org charts at most enterprises, it will be profound. But in terms of software development, our platform runs great in a cloud or runs great right here on the ground. We leave it up to our customers to figure out where they want to deploy."
Jun 18, 2013 |
The world's largest supercomputers, like Tianhe-2, are great at traditional, compute-intensive HPC workloads, such as simulating atomic decay or modeling tornados. But data-intensive applications--such as mining big data sets for connections--is a different sort of workload, and runs best on a different sort of computer.
Jun 18, 2013 |
Researchers are finding innovative uses for Gordon, the 285 teraflop supercomputer housed at the San Diego Supercomputer Center (SDSC) that has a unique Flash-based storage system. Since going online, researchers have put the incredibly fast I/O to use on a wide variety of workloads, ranging from chemistry to political science.
Jun 17, 2013 |
The advent of low-power mobile processors and cloud delivery models is changing the economics of computing. But just as an economy car is good at different things than a full size truck, an HPC workload still has certain computing demands that neither the fastest smartphone nor the most elastic cloud cluster can fulfill.
Jun 14, 2013 |
For all the progress we've made in IT over the last 50 years, there's one area of life that has steadfastly eluded the grasp of computers: understanding human language. Now, researchers at the Texas Advanced Computing Center (TACC) are utilizing a Hadoop cluster on its Longhorn supercomputer to move the state of the art of language processing a little bit further.
Jun 13, 2013 |
Titan, the Cray XK7 at the Oak Ridge National Lab that debuted last fall as the fastest supercomputer in the world with 17.59 petaflops of sustained computing power, will rely on its previous LINPACK test for the upcoming edition of the Top 500 list.
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/15/2013 | Bull | “50% of HPC users say their largest jobs scale to 120 cores or less.” How about yours? Are your codes ready to take advantage of today’s and tomorrow’s ultra-parallel HPC systems? Download this White Paper by Analysts Intersect360 Research to see what Bull and Intel’s Center for Excellence in Parallel Programming can do for your codes.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?
Join our webinar to learn how IT managers can migrate to a more resilient, flexible and scalable solution that grows with the data center. Mellanox VMS is future-proof, efficient and brings significant CAPEX and OPEX savings. The VMS is available today.