Search results matching tag 'HPC'

SC09 Conference starts Monday in Portland

alanna.dwyer@hp.com — Sat, 14 Nov 2009 01:00:00 GMT

Several of us are already in Portland, in advance of the HPC conference, to attend HP-CAST. HP-CAST is a user group focused on HPC, whose fall meeting is usually held right before the industry-wide SC event. This 13^th Group Meeting just got kicked off this morning with a preview on next week’s technology announcements from HP, and updates from some HP-CAST members on site deployments and key research activities and objectives. Look for some announcements next week which will highlight some new sites.

SC09 is the major HPC conference in the world – this year it is in Portland, Oregon. Portland is known for its environmental activism – making it appropriate site for the meeting this year, as customers and vendors are all driving to improve power utilization while driving performance. HP booth’s will showcase datacenter power and efficiency technologies. There’s a demo that highlights the SL line – the ProLiant SL2x170z just recorded the best SPECpower numbers for any server. Other systems on display include latest ProLiant G6 BladeSystems, the ProLiant DL1000, and the new scalable storage system that was introduced last week (the x9000 Network File System). Accelerators and graphics-enabled servers will also be shown. Cluster tools from HP and our partners will be shown – one of growing interest is ‘hybrid clusters of Linux and Windows’ – in which one can quickly change the operating environment to Linux or Windows stacks, to support specific workloads. The specific implementation we have is using Moab HPC Adaptive Suite to direct provisioning and scheduling based on workload, with HP CMU managing the Linux stacks, and Windows HPC Server as the Windows base. Drop by if you are in the area!

HP saw the wheels of recovery turning last week at HP's 20th Annual CAE Symposium....in Detroit!

alanna.dwyer@hp.com — Fri, 09 Oct 2009 15:29:00 GMT

I got the following report from Christine Fronczak , our WW Computer-Aided Engineering Business Manager, highlighting observations from the recent HP CAE Symposium (see prior blog), and thought it would be intersesting to share with you all. Thanks, Christine, for permision to post! Note that presentations from the Symposium are available for download at the site www.hp.com/go/cae.

"It was a great honor for me to host HP’s 20^th Annual CAE Symposium last week in Detroit. For our 20^th anniversary event, we chose to return to Detroit… the locale of so many successful CAE events with anchored by several key CAE automotive customers. It’s been quite a climb for these customers during these many months of a difficult economy, but after meeting with customers at the event, I was convinced that Detroit has the perseverance and drive to recover and return to a leading center of innovative product design and development. The discussions that I had with key customers confirmed their belief in CAE simulations and recent HP CAE investments by these customers prove that they realize the criticality of CAE to ensure their growth and success for the US automotive industry. One customer told me that even though times are difficult, cuts in CAE simulations would be ill advised because lack of investment today would result in lack of competitive product offering when the economy fully recovers.

I also learned of interesting and innovative new ways CAE simulations are being used. The footprint of applications for CAE is expanding and is now treading in product designs and other areas that were previously sacred to PLM. Customers are also looking for alternate deployment models for CAE, such as the use of flexible computing resources through hosting companies, such as HP’s Flexible Computing Services. Yes, we’re talking about CAE as a Service. Customers see this as a “safe” way to expand available resources during peak usage. We are not talking your grandfather’s CAE!

Our day began with very compelling keynotes. HP’s Ed Turkel kicked things off with a brief history of CAE in HPC. The audience highly valued Keith Meintje’s keynote (CPDA) “The Rise of Simulation—The History, and Future, of CAE” which showed the vast improvements made in this technology area but also revealed the opportunities for growth that are yet to come. All of the attendees are asking for a copy of Keith’s slides!

Also featured on our keynote agenda were presentations from our premiere event sponsors, Intel and Microsoft.

Intel’s performance leadership today with their Nehalem chipset was shown but they also revealed a roadmap with further enhancements for next year. That roadmap got this CAE crowd pretty excited….nothing like a good boost in performance to improve your CAE simulations!

Microsoft is also investing heavily in HPC and, in particular, the CAE segment. The industry’s use of HPC Windows 2008 is growing and Microsoft provided a valuable proof point by sharing their limelight on stage with CAE customer Honeywell Aero. Christopher Chang (Honeywell) shared how his team, original Linux experts, grew to love and respect the Microsoft Windows environment because it provided reductions in cycle times (11 days to 3 days!) and decreased their overall time to solution. Honeywell is growing (not cutting) their budgets for CAE simulations.

The afternoon was filled with three simultaneous tracks (2 with CAE ISVs and the 3^rd with HP/EDS solutions & services). All were heavily attended and we saw customers lingering in discussions well after the 6:00 p.m. ice cream social had ended.

This year’s event has got me really fired up about the CAE market –and thinking about all kinds of exciting new applications for the category to help our customers shorten development cycles and time to market, saving them money in the process.

For further info about the event and to see the presentations, please visit our website at www.hp.com/go/cae"

20th HP CAE Symposium hits Motor City to talk advances in computational engineering

alanna.dwyer@hp.com — Tue, 29 Sep 2009 21:46:00 GMT

This week, HP’s high-performance computing team hits Detroit for our 20^th HP CAE Symposium. Bringing together customers, partnering and CAE thought leaders, this year’s event will focus on the power of computer-aided engineering to speed product development and time-to-market, even –or perhaps especially –when the economy goes haywire.

We’re really pleased to be holding this year’s event in Motor City, amongst so many of our key customers in the extremely competitive automotive segment in which CAE is critical.

Here are a few of the highlights of this year’s agenda:

GE alum and CAE analyst, Keith Meintjes, currently research director, CAE, at CDPA will keynote on “The Rise of Simulation- The History, and Future of CAE”
HP’s own Ed Turkel, manager, business development, Scalable Computing & Infrastructure, will talk “Trends in HPC and CAE”
HP partners Intel and Microsoft will keynote at this year’s event.
Partner solution showcase will feature technology and solution demos from the following: Acusim Software, Altair Engineering, ANSYS, CD-adapco, CEO, Cradle, ESI Group, Exa, Intel, Livermore Software Technology Corp (LSTC), Mellanox, Microsoft, MSC Software, Platform Computing, QLogic, Siemens, SIMULIA, TASS and Voltaire

More information on the event: www.hp.com/go/caesymposium2009

More information on HP in CAE: www.hp.com/go/cae

Desktop Engineering magazine is the exclusive media sponsor of this year’s CAE Symposium: http://www.deskeng.com/

HP Tops Top500 Again!

alanna.dwyer@hp.com — Wed, 15 Jul 2009 13:42:00 GMT

The new Top500 list was released recently, and HP (and BladeSystems) again tops the list as the preferred technology choice by this extremely knowledgeable and demanding community of users! What's the Top500? It's a listing of the world's most powerful supercomputing systems, based on the Linpack benchmark, and is compiled twice a year. HP systems power 212 of these 500 sites (42%). Of those, 207 sites deploy HP c-Class BladeSystems, making it the most popular and fastest growing architecture on the list. Notable trends include: the continued growth in InfiniBand as high performance interconnect, especially for the top tier in the Top500; demonstration of Windows HPC scalability, with that technology being used at the #15 site (in Shanghai); and new countries gaining ground in the top tier, such as Saudi Arabia with the #14 spot.

Interconnect technologies: InfiniBand, 1G Ethernet, 10G Ethernet

guodong.zhang@hp.com — Thu, 02 Jul 2009 16:40:00 GMT

Interconnect is an integral part of any scale-out solutions. The question often asked is something like "which interconnect should I use for my next scale-out project?" Well, a simple answer is "it depends". For many applications in the high performance computing (HPC) area, their performance highly depends on the latency and bandwidth of the interconnect fabric -- majority of MPI-based parallel applications are in this category. InfiniBand has been proven to be the best choice for these applications. In the latest released Top500 (June 2009), I counted 151 entries with InfiniBand as the interconnect, increased from 124 6 months ago. Another interesting observation is about what server platforms were used to build these clusters -- I found that the HP BladeSystem c-Class is the most popular server platform with 37 entries, followed by IBM's pSeries with 17 entries. In the last couple of years, we also see more and more cases in commercial applications where InfiniBand is being used as a transport interconnect to meet the low latency and scalability requirements, examples including HP ExaData storage server, Market data systems deployed in Financial Services Industry, and applications in rendering applications.

Not every application needs InfiniBand. There are workloads, include some HPC workloads, that their performance is not sensitive to the latency of the underline interconnect. For these application, Gigabit Ethernet provides a very cost effective solution. However, we see increased bandwidth requirement in many of these cases, partially driven by the adoption of server virtualization, or simply to match higher I/O requirement driven by significant performance increase of the multi-core servers. 10Gigabit Ethernet is becoming a natural choice to meet this higher bandwidth requirement, yet, allow people to stay within the technology they are familiar with.

So, to summarize, the decision on the choice of interconnect should be based on the performance requirements of applications: for majority of MPI parallel applications as well as those that require low latency and high bandwidth, InfiniBand is becoming the interconnect of choice. And InfiniBand is also a technology that is very easy to use and manage.

HPTF 2009: "What happens in Vegas doesn't have to stay in Vegas!!" Stay connected.

Kristie Popp — Sun, 14 Jun 2009 14:54:00 GMT

Traveling to the show? Share with us your experiences. Can't travel? Stay tuned to this web site for the latest info.

At Tech Forum (June 16 - 18 in Las Vegas) HP will showcase its premier enterprise computing solutions. The event will feature daily keynote speakers, presentations and booth demos. If travel restrictions are preventing you to attend in person, you still have the opportunity participate. Via a dedicated website that includes a wide range of tools, HP will provide access to summary videos of selected keynote presentations and demos. See this year's choices of topics at the event & expo.

"What happens in Vegas doesn't have to stay in Vegas!!" Check www.hp.com/go/HPTF-buzz for all the info on the HP Tech Forum 2009 for servers, storage and software. Share with us what you are seeing and experiencing on Twitter, Facebook, Blogs, Slideshare, upload your photos to Flicker and videos to YouTube.

More blog info coming from the show including lists of the "top threes" from the event. Please also feel free to comment on the site www.hp.com/go/HPTF-buzz to help us give you the info you want. Likes, dislikes?

Less Skin in the Game

alanna.dwyer@hp.com — Wed, 10 Jun 2009 14:52:00 GMT

Today HP announced the new ExSO portfolio. The products are designed for Extreme Scale Out (duh) but are quite applicable for HPC and clusters. After all, the typical HPC user wants highest performance per dollar, and per watt. Yes, these systems are ideal for big Web 2.0 and cloud data centers, but the economics make a lot of sense for mere mortals with 100+ nodes. As the announcement explains, the new ProLiant SL family uses a ”skinless” systems architecture that replaces the traditional chassis and rack form factors with an extremely lightweight rail and tray design.” Multiple nodes share power units and fans, which deliver better power utilization and cooling than possible in traditional 1U servers, as we found with the HP BladeSystem. But also the skinless design means less metal to retain heat, and also less weight. Each SL6000 chassis holds 2 trays. There are currently 3 basic servers available (you can mix and match in a standard rack) – there are the 2 nodes in 1U tray for compute intense apps, a large memory node in 1 U tray for memory intense apps, and a node with up to 6 disks. The energy efficiency is further enabled with the Intelligent ExSO Center including the new Data Center Environmental Edge that provides a visual map of the data center environmental variables.

SLURM: a Simple Resource Manager no more!

uathb — Mon, 08 Jun 2009 20:31:00 GMT

SLURM, the "Highly Scalable Resource Manager" from Lawrence Livermore National Laboratory, has come a long way. When we started using it for some of the products in our Unified Cluster Portfolio, SLURM had only a simple FIFO scheduler, no job accounting, and the ability to support perhaps a thousand nodes.

SLURM also provided a very clean architecture which allowed HP to contribute the first versions of job accounting for Linux clusters, support for multi-threaded and hyperthreaded architectures, complex job scheduling such as gang scheduling, and fine-grained allocation of system resources ("consumable resources").

SLURM version 2.0 has just been released, and what a powerhouse it is! Heterogeneous clusters, up to 65,000 nodes, resource limits, and job prioritization. Moe Jetty and Danny Auble, the primary authors of SLURM, discuss it on this podcast.

My compliments to the entire team!

HPC at HP Tech Forum (June 15-18)

alanna.dwyer@hp.com — Fri, 05 Jun 2009 12:38:00 GMT

Posts from Mark and Kent in this blog have covered topics that we’ll be covering in two weeks at the HP Technology Forum, in Las Vegas. Mark is delivering two talks on Cluster Monitoring, including a deep dive on Collectl, the open source project led by Mark. Accelerators are a featured talk, and will be on display in demos in the Expo area. There are also talks on new solutions for Extreme Scale-Out, BladeSystems in HPC, Cluster Management, Scalable Storage, the POD (Performance-Optimized Datacenter), and several others. A POD container will be at the show, and several cluster demos will be running in the Expo. Go to www.hptechnologyforum.com for more information. Hope to see some of you there.

"To Boldy Go" - HP Clusters enable spacecraft design

alanna.dwyer@hp.com — Thu, 28 May 2009 18:32:00 GMT

So it's not brain surgery, but it is rocket science!

See article in Supercomputing Online.

http://supercomputingonline.com/index.php/2009050816472/latest/orbital-sciences-corporation-uses-hp-supercomputer-to-design-the-next-generation-of-spacecraft.html