Visit additional Tabor Communication Publications
April 08, 2010
Microsoft is releasing the second beta version of its latest HPC server platform -- Windows HPC Server 2008 R2. The first beta was released at SC'09 last November and in the interim the company has been collecting customer feedback. All of this is expected to culminate in a formal release later this year.
Microsoft, of course, is attempting to establish a foothold in the wonderful world of high performance computing -- a tough undertaking in this Linux-dominated and open-source-loving IT segment. That said, the additions to this latest beta version of the Microsoft's HPC platform include some compelling features that make the most of the company's vast Windows ecosystem and market reach.
Before I talk about the new features, it's worth noting that Microsoft believes it has reached parity with Linux from the performance standpoint. Vince Mendillo, senior director of Microsoft's High Performance Computing Group, told me they're continuing to tune the software stack as each beta rolls out, and they're now at the point where they feel performance is a non-issue. According to him, using Linpack or even real-world benchmarks like ANSYS Fluent 12.0, Windows HPC is now as fast or faster than a Linux implementation. But Microsoft's real strengths are the Windows development tools and its ecosystem dominance, and it is here that the company has focused its efforts.
For example, Microsoft has added the ability to aggregate Windows 7 workstations into an HPC cluster. Essentially, each workstation is monitored and managed as an ad-hoc compute node within a traditional cluster. Capabilities like time-of-day scheduling and shutting down of preemptive jobs are included so that the machine can be made available to a real live person when required. The common scenario is one where an organization has a small cluster made up of say dozens of servers, along with perhaps hundreds of Windows 7 PCs sitting idle at night.
Since these workstations are likely to be hooked up to the LAN via a high-latency Ethernet connection, the type of HPC workloads that can be run on them is somewhat more limited than that of a typical InfiniBand-connected cluster. In theory, any highly parallel but loosely coupled application would be suitable. If you have such a setup, you can essentially add a night-time cluster for free (well, at least for the cost of the power to run it).
The next big enhancement is the integration with Visual Studio 2010, which is due to be launched next week. This new version of Visual Studio contains lots of support for creating, testing and deploying parallel apps. It includes the hooks for Microsoft's MPI, SOA-type execution, and new services for Excel 2010. Mendillo characterized the integration between Visual Studio and Windows HPC Server as a "big breakthrough moment." He may be right. Visual Studio has over 7.5 million users today, which is the largest user base for any software development suite. Given that, the ability to program from the client across the cluster is a big deal.
The Excel tool has received special attention since it has become a fairly widespread tool for HPC-types -- scientists, engineers, financial quants, and the like. Basically, Microsoft has layered HPC services on top of vanilla Excel so that users can distribute computations across a cluster. Mendillo says there are a lot of Excel applications that are suited to this type of parallelization. One example is a life insurance actuarial calculation. On a workstation, a typical such calculation could take 14 hours. Mendillo claims distributing that work across a small cluster -- say 16 to 32-nodes -- cuts that time down to two and a half minutes. Microsoft is partnering with integrators Wipro, Infusion and Grid Dynamics to help customers migrate their Excel workbooks and develop new parallelized apps.
Along these same lines comes Parallel Nsight, NVIDIA's software development toolset for GPU computing that is offered as a Visual Studio plugin. Microsoft has been working with NVIDIA on this for a while, but the software maker's big contribution is low-level GPU support on the server side so that everything works seamlessly end-to-end. The whole package allows programmers to develop, test, and run GPU-accelerated applications, all with the Windows universe. This has the makings of a nice synergy for both vendors as it allows NVIDIA to piggy-back on the ubiquity of Visual Studio and gives Microsoft a GPU development capability alongside its CPU toolset.
The last piece of the puzzle for Windows in HPC is Linux interoperability. Although some dual-boot operating system capability has been there in the past, the beta release adds compatibility with mixed-OS cluster management platforms from Adaptive Computing, Clustercorp and Platform Computing. This allows the system admin to statically boot specific nodes with Linux or Windows. In the case of Platform Computing, dynamic OS provisioning can be supported via the company's ISF Adaptive Cluster platform. The whole idea here is to be able to serve Windows and Linux HPC apps on the same hardware, and to do so as transparently as possible. This is a must-have feature for Microsoft, inasmuch as HPC customers have come to expect Linux as the default OS.
If any of this arouses your interest, Microsoft is making the beta available for downloading on its Windows HPC Server trial site. There could be another beta round says Mendillo, or they may go straight to the final release. In either case, he expects the Windows HPC Server 2008 R2 product to be shipping later in the fall.
Posted by Michael Feldman - April 08, 2010 @ 6:31 PM, Pacific Daylight Time
Michael Feldman is the editor of HPCwire.
No Recent Blog Comments
During a conversation this week with Cray CEO, Peter Ungaro, we learned that the company has managed to extend its reach into the enterprise HPC market quite dramatically--at least in supercomputing business terms. With steady growth into these markets, however, the focus on hardware versus the software side of certain problems for such users is....
Contributing commentator, Andrew Jones, offers a break in the news cycle with an assessment of what the national "size matters" contest means for the U.S. and other nations...
Today at the International Supercomputing Conference in Leipzing, Germany, Jack Dongarra presented on a proposed benchmark that could carry a bit more weight than its older Linpack companion. The high performance conjugate gradient (HPCG) concept takes into account new architectures for new applications, while shedding the floating point....
Jun 19, 2013 |
Supercomputer architectures have evolved considerably over the last 20 years, particularly in the number of processors that are linked together. One aspect of HPC architecture that hasn't changed is the MPI programming model.
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.
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.