Entries tagged with high performance computing - Channel 9

Dan Reed: On the ManyCore Future and Parallelism in the Sky

Charles — Thu, 10 Apr 2008 19:42:12 GMT

Dan Reed is Microsoft's Director of Scalable/Multi-Core Systems Research and head of the recently formed Universal Parallel Computing Research Centers (UPCRC): one at the University of California at Berkeley (UC-Berkeley) and a second at the University of Illinois at Urbana-Champaign (UIUC). Since we've been here) but part of the same wide-angle general purpose solution to the complex (and arguably fractal) general problem that spans microblips in DRAM to massive data centers.

Certainly the computation Cloud of the future must not only be scalable and highly performant, but also adaptive and homeostatic in how it reacts to frequent perturbation.

What are some of the challenges on the server side with respect to concurrent processing and massive scalability? Clustered server computing environments have traditionally been very good at parallel computation (compared to the general purpose client) so what's Dan and Microsoft working on to ensure our Cloud scales to ManyCore? Is machine learning being incorporated into clustered computing software adaptation and evolution?

Dan has a very interesting biography:

"Previously, I was the founding director of the Renaissiance Computing Institute (RENCI) at the University of North Carolina, the Chancellor's Eminent Professor, and Senior Advisor for Strategy and Innovation. Before that, I was head of the Department of Computer Science, Edward William and Jane Marr Gutgsell Professor, and Director of the National Center for Supercomputing Applications (NCSA) at the University of Illinois.
I am also a member of the President's Council of Advisors on Science and Technology (PCAST) and chair of the Computing Research Association (CRA)"

Dan was the head of CS at Illinois during the birth of the web browser Mosaic which changed the way people interact with the Internet forever... We talk about where the web is today (including browsers) versus what Mosaic enabled when it arrived.

Enjoy. This is another great discussion with a supercomputing stalwart whose main focus these days is on ensuring we are prepared for the highly parallel future of general purpose computation in the sky.

Low res file here.

Intro to Cluster Diagnostics and Node Deployment

philpenn — Thu, 06 Mar 2008 07:00:44 GMT

Cathy is a Senior PM on the HPC Server 2008 product team. Follow along as Cathy demonstrates new diagnostics and deployment features.

Windows HPC Server 2008 is available in a Beta1 release as of November, 2007, with Beta2 scheduled for release in Spring, 2008. Obtain the beta bits at http://connect.microsoft.com. Please provide feedback using the feedback tool on the Connect web site.

For additional information, see http://WindowsHPC.net and http://edge.technet.com/Tags/HPC/.

Five Minute Intro to the HPC Server 2008 Management Console

philpenn — Thu, 06 Mar 2008 05:20:59 GMT

Cathy is a Senior PM on the HPC Server 2008 product team. Follow along as Cathy demonstrates the new Cluster Management Console.

Windows HPC Server 2008 is available in a Beta1 release as of November, 2007, with Beta2 scheduled for release in Spring 2008. Obtain the beta bits at http://connect.microsoft.com. Please provide feedback using the feedback tool on the Connect web site.

For additional information, see http://WindowsHPC.net.

Burton Smith: On General Purpose Super Computing and the History and Future of Parallelism

Charles — Wed, 13 Feb 2008 16:34:00 GMT

Burton Smith is a Technical Fellow at Microsoft who thinks about ways in which our platform needs to be structured to support general purpose computers that will soon have clustered super computer processing power as we move closer to manycore everywhere (not too far off into the future...). Burton is a parallel computing expert, an industry thought leader in high performance, massively parallel distributed (aka super) computing. Winner of the Seymour Cray Computer Engineering Award, Burton knows a thing or two about how to architect and implement software systems that can succeed in the Age of Manycore.

This is a long and great conversation, unedited of course. You'll want to make some time for this and listen carefully to what Burton says. This is a very important general introduction to parallelism and high performance computing. As always, we can't talk about super computing without addressing program language evolution in the context of manycore (you've seen this quite a bit on C9 over the years). We cover a lot of ground here including Burton's insights into functional programming, transactions, compatability, shared mutable state, operating systems, technical redunancy and the role of Technical Fellows in the post-Bill era.

Enjoy this great introduction to parallelism and the future of our platform technologies and tools as we head into the age of manycore. This is the first in a series of several interviews covering parallel computing and Microsoft's Parallel Computing Platform technologies, specifically.

Low res file for the bandwidth-challenged.

New OGF Web Service Interface with Windows HPC Server 2008

philpenn — Wed, 02 Jan 2008 21:29:19 GMT

Steven is a member of the Windows HPC Server 2008 product team responsible for design and development of the new Open Grid Forum compatible Web Service interface. This new interface enables client HPC applications to interact with the compute cluster job scheduler in a vendor-neutral manner.

Learn more about this and other new features via the community website at http://WindowsHPC.net.

The video is a bit long for Channel9 (~30 min). You may wish to download the entire WMV file, open it via the Windows Media Player, and select the fast play speed option, "Ctrl+Shift+G".

New performance enhancements for MPI applications running on HPC Server 2008

philpenn — Tue, 18 Dec 2007 04:11:11 GMT

Eric is a member of the HPC Server product team. Listen to Eric describe the new Network Direct RDMA interface and related performance observations. Microsoft's MPI library is the first user-mode application that is also Network Direct aware. Resultant benefits include very low latency metrics.

Learn more at http://blogs.msdn.com/philpenn and http://WindowsHPC.net.

Getting Started with Windows HPC Server 2008

philpenn — Thu, 15 Nov 2007 21:35:23 GMT

Beta1 of Windows HPC Server 2008 (aka Compute Cluster Server) has just been released (Nov 2007). The product consists of a unique Windows Server 2008 edition intended for deployment in compute intensive "Beowulf type" High Performance Computing clusters. Included are deployment, operations and scheduling services as well as libraries for development of "cluster aware" applications.

HPC is just one option for parallel computing solutions and specifically intended for localized high bandwidth, low-latency, deployments.

Please also see the following sites:

1. Windows HPC Community: http://WindowsHPC.net

2. My blog: http://blogs.msdn.com/philpenn

3. Product web: http://www.microsoft.com/hpc

geekSpeak Recording: Programming for High-Performance Computing Environments with Dariusz Parys and

glengo — Fri, 26 Oct 2007 23:22:37 GMT

We're always trying to dig up esoteric subjects for geekSpeak, and here's one that is intriguing - High Performance Computing. Once strictly the domain of universities and the like, HPC is quickly becoming a cornerstone of business analysis, engineering and number crunching in general.

This geekSpeak features several Microsoft folks from Germany, including Developer Evangelist Dariusz Parys, Platform Strategy Manager Christian Binder and infrastructure specialist Steffen Krause.

These fine gentlemen get us all familiar with key aspects of developing and managing applications for a high-performance computing environment. Steffen gives a great overview of a typical infrastructure layout, which includes Windows Compute Cluster Server 2003. We get some great tips for handling multi-threading through the use of OpenMP, as well as an overview of using the Message Passing Interface (MPI) to handle the communication to the cluster.

Steffen shares with us some information around using Excel services on Sharepoint actually installed on a cluster to provide some serious horsepower for server-based calculations. And we hear about a specific HPC scenario from Sorin Serban, with Visual Numerics, involving some amazingly complex algorithms.

All in all, a fascinating geekSpeak on the possiblities that high-performance computing offers.