One of the areas where High Performance Computing (HPC) gets utilized is simulation.
Luke Mihelcic, SIM Squad member and Technical Marketing Manager, and Sam Murgie, SIM Squad member and Software Development Manager talk about the use of High Performance Computing (HPC) in Simulation.
During the ISC Cloud’11 event Wolfgang Gentzsch lead a panel discussion with Kenichi Miura, Per Oster and Tom-Michael Thamm to offer insights into advantages and challenges of running high performance computing (HPC) applications in the cloud.
The panel members listed a few real world challenges of running HPC applications in the cloud:
Cloud offers have trouble matching performance of purpose built HPC clusters
The cost savings aren’t guaranteed – especially due to data transfer/storage costs
Data transfer speeds can be prohibitive
Security is a concern in public clouds
The panel also comments that cloud technologies should be complimentary and not a replacement to the HPC.
Small businesses need for access to HPC resources is increasing every where in the world including Europe. To remain competitive in an market space where competitors are turning to digitized manufacturing methods, computer aided design and simulation access to high performance computing resources is becoming a necessity.
In Europe a number of initiatives are underway to address this need. If you are a small business looking for guidance on access to high performance computing resources (HPC) you may find the following contacts useful:
PlanetHPC which is a Support Action funded by the EU’s 7th Framework Programme in November 2009
PRACE (Partnership for Advanced Computing in Europe) which is an international non-profit association
Similar initiatives are active in North America as well. Read more at: HPC for small businesses
Do you know of other initiatives? Drop us a note below or email us at: support@stackthem.com
Small businesses need for access to HPC resources is increasing. To remain competitive in an market space where competitors are turning to digitized manufacturing methods, computer aided design and simulation access to high performance computing resources is becoming a necessity.
In North America a number of initiatives are underway to address this need. If you are a small business looking for guidance on access to high performance computing resources (HPC) you may find the following contacts useful:
If you are looking for HPC resources in Europe, please see: HPC for Europian Small Businesses
Do you know of other initiatives? Drop us a note below or email us at: support@stackthem.com
Guess what percentage of products are expected to be manufactured via digitized methods by 2020.
If you said 50% you aren’t even close. Try 98% according to National Center for Manufacturing Sciences (NCMS).
With digitization of manufacturing processes and intense global competition comes the need for computing resources that small and medium size business can utilize.
We are not talking about desktop computers or even more powerful workstations generally used for computer aided design applications. Small and medium manufacturers will need access to high performance computing resources to power more advanced applications such as simulation.
Will US manufacturers be able to get access to HPC resources and how? NCMS and few other organizations are working on addressing this very need. More on that in a future post.
StackThem is working on helping manufactures gain access to HPC resources as well. If interested check out www.stackthem.com.
Cloud computing is changing the world of IT and job titles aren’t going to escape from its transformation either.
In a recent article Forbes published a list of cloudy job titles that are cropping up.
Any Virtualization & Cloud Computing Subject Matter Experts out there?
Is this because cloud is new and hasn’t become mainstream yet? If the cloud is a mainstream concept in IT, shouldn’t a “Cloud Computing Architect” be simply called an “Architect”? We don’t have Internet Architects around do we?
Original article:
http://www.forbes.com/sites/joemckendrick/2011/12/26/cloud-computing-is-changing-many-job-descriptions/
Ran into this youtube video about a supercomputer for under $500,000.
Nagasaki University Assistant Professor Tsuyoshi Hamada talks about how it was build. Fun to watch:
Hamada won Gordon Bell prize for price/performance in 2009 and was nominated again in 2010.
Will supercomputers become more readily available to companies other than the few fortunate (or wealthy) enough to own one?
In a recent post Tyler O’Neal explains:
“For now, high-tech universities, government research labs, and major corporations use supercomputers on a regular basis, major organizations from the MIT to the Pentagon to Hollywood movie companies.”
Amazon’s EC2 HPC is one way of getting your hands on one for sure, but there is another way.
In the same article Tyler introduces HokieSpeed with a quote from Wu Feng, associate professor at the Virginia Tech College:
“HokieSpeed is a versatile heterogeneous supercomputing instrument, where each compute node consists of energy-efficient central-processing units and high-end graphics-processing units.”
Virginia Tech College is planning to make HokieSpeed available for researchers at the College and the industry. As more and more supercomputers are built they will become more available to the masses. Wu Feng summarizes:
“The next frontier is to take high-performance computing, in particular supercomputers such as HokieSpeed, and personalize it for the masses.”
Isn’t there a missing link here? How will the masses go about locating these supercomputer centers popping up all over the country?
StackThem provides a meeting place for supercomputer owners and researchers who need access to them. Read more at: www.stackthem.com
Original article:
http://www.supercomputingonline.com/latest-news2/23960-virginia-techs-wu-feng-unveils-hokiespeed-a-new-powerful-supercomputer-for-the-masses
Enterprise resources are frequently procured to serve specific purposes such as holding the data of a division or crunching the data for a project.
Over time, the resources of the enterprise gets locked into silos, with minimal central control and visibility. This is a leading driver of low utilization rates many high performance clusters suffer from.
IBM is seeing an opportunity in providing management tools to break down the silos and expose the resources in a centrally manageable fashion.
In a recent post Eric Doyle (@ITscribe) writes:
“By providing the tools required to manage an HPC infrastructure using cloud computing principles, the silos can be dispelled and resources can be allocated on a project-by-project basis. This maximises the operational efficiency of existing computational clusters by creating a pool of resources.”
Original Article:
http://www.techweekeurope.co.uk/news/ibm-brings-supercomputing-into-the-cloud-age-31464
This is a great initiative. But how will the industry break the silos that cross corporate boundaries?
StackThem has set out to do just that. For more information visit www.stackthem.com.
2011 shaped out to be the year when public cloud providers captured the attention and wallet share of many IT departments. The speed with which applications are moving to the public clouds (also known as infrastructure as a service) is staggering.
But will public clouds replace in house HPC (high performance computing clusters)?
In a paper titled “Smart IBM Solutions for High Performance Engineering Clouds” Srini Chari (@SriniChari), of Cabot Partners Group, Inc sums it up very well:
“…the main problem with running HPC or engineering tasks on conventional clouds is that conventional clouds are geared toward supporting general-purpose applications and services — short transactional workloads such as Web applications and database tasks. HPC tasks, on the other hand, are mostly complex, long-running algorithms processed in parallel, with the result of one task not dependent on the outcome of another. Processing threads are brought together at the end of the activity.”
If public clouds isn’t such a great alternative to in house HPC, will owners of high performance computing clusters continue to suffer from the classic problem of capacity utilization?
Not if they can find a way to rent out their idle HPC capacity to others who are willing to pay for it. StackThem offers a solution: www.stackthem.com
stackthem 