Archive

Archive for the ‘Cloud Computing’ Category

Leveraging Cloud Computing in CAE Product Development Environments (Part II)

March 29th, 2011 No comments

Client Risks and Limitations

Security Concerns

Will data be secure in cloud based product development?

While it is important to that vendors are selected with care, cloud based processing and storage of sensitive data has become common within Fortune 500 companies.[1] Standards and certifications exist to ensure that no unauthorized access occurs and sensitive data is kept secure.[i]

Scalability Concerns

How scalable is computational processing?

Disciplinary solvers for specific engineering disciplines will always have natural limitations to scalability.  However, many current processes are not yet approaching diminishing returns and a variety of opportunities exist to exploit coarse-grain and medium-grain parallelization immediately.  Furthermore, integration across functional disciplines will provide further opportunities for concurrent processing. 

Integration Concerns

How will disciplinary software be integrated?

Software providers do need to make appropriate adjustments to fully integrate with cloud based platforms.  Current vendors provide various levels of integration opportunities through API interfaces.  Incentives are aligned for software vendors to continue the development of these interfaces to integrate closely into cloud based platforms.  Additionally, leading CAE vendors indicate a strong willingness to participate in further development opportunities.

Immediate Opportunities

The following opportunities can be explored to move toward a cloud based product development vision.  Each opportunity can be examined in the form of a pilot project and be implemented in a low risk controlled environment within which benchmarking tests can be completed to verify and quantify the extent to which clients can benefit from cloud based product development.

Cloud Based High-Performance Computing (complementary to existing services)

Cloud based HPC will allow clients to immediately start capturing the benefits of additional computing resources in a controlled environment.  Benchmarking tests for functional disciplines will allow accurate quantification of the benefits that cloud based high-performance computing can provide within various product development environment verticals.  Integrating third-party cloud computing services with the existing CAE tool suite will allow clients to immediately access additional computing power when needed.  Providing these capabilities will eliminate barriers to entry for both existing and new clients to experience the benefits of cloud based HPC.

Virtualization experts[ii],[iii],[iv] and case studies[2],[v],[vi] have verified that cloud based high-performance computing is beneficial and viable from both an academic and practical perspective.  Leading automotive and aerospace clients have demonstrated a strong interest to evaluate cloud based HPC opportunities.

Cloud Based Integration Platform (complementary to cloud based HPC)

Developing a cloud based platform to integrate disciplinary software and other third-party product development tools provides the opportunity for clients to explore the product performance benefits that can be gained from integration across multiple functional silos within an organization regardless of which software providers are used.  Benchmarking tests with clients will demonstrate the level at which current software and third-party tools can be integrated through using existing application programming interfaces (API) while using cloud based HPC for computation power.

Existing software integration platforms lack both cloud based capabilities and a deep enough integration with third-party independent software vendors to provide the full benefits that an integrated platform could provide.  Developer tools and customized implementation of a cloud based integration platform will be able to provide a deeper integration into third-party software and eliminate the computational bottlenecks that make leading software integration platforms impractical for large-scale production product development environments.

Independent software vendors (ISV) and clients in relevant industry verticals have both shown interest to participate in and deploy cloud based integration platforms.  There are numerous incentives and benefits to CAE software providers, third-party ISV’s, and clients to participate in a cloud based product development platform.

A Call to Action

Moving toward cloud based product development environment will provide numerous benefits to the product development processes of clients that will reduce development timelines, improve product performance, and reduce costs.  Cloud computing technology has matured far enough that the benefits far outweigh the risks associated with the proposed developments.  Leading software providers and industry pundits agree that the transition to cloud based computing and smart integration of software tools will be two main sources of value creation in computer aided engineering tools in the short to medium term.[vii]

“[Organizations] run [cloud based HPC] applications such as mapping genomes for scientific research, simulating aerospace and automotive designs for engineering activities, mining data for business intelligence and many other use cases.”[viii] Recent case studies have addressed the risks and demonstrated magnitudes of improvements on performance and drastic reductions in cost which easily make the case for cloud based HPC on a standalone basis, exclusive of the significant additional potential benefits from integration across functional disciplines in a cloud based product development platform.

A proactive investment to explore the aforementioned benefits in cloud based product development is likely to provide invaluable returns for clients in the form of sustainable competitive advantage over the longer term.  With swift action, clients will quickly be able to start leveraging the benefits of cloud computing technology, reducing the time and cost to innovate and improving the performance of their products.


[1] Salesforce.com illustrates the willingness of leading companies to trust the security of cloud based services with some of their most sensitive data

[2]Josh Simons mentions (at ISC Cloud ’10)  a successful case study in financial services where significant benefits were realized from moving to cloud based HPC across multiple disciplines, http://communities.vmware.com/docs/DOC-13947


[i] “Amazon Web Services Overview of Security Processes”, Amazon.com, http://media.amazonwebservices.com/pdf/AWS_Security_Whitepaper.pdf, August 2010

[ii] “Why Virtualize HPC?”, Josh Simons, http://communities.vmware.com/community/cto/high-performance/blog/2010/08/25/why-virtualize-hpc, accessed February 2011

[iii] “Virtualization for High Performance Computing”, Mark F. Mergen, Volkmar Uhlig, Orran Krieger, Jimi Xenidis, IBM Research Center, ACM SIGOPS Operating Systems Review, Volume 40 Issue 2, April 2006

[iv] “A case for high performance computing with virtual machines”, Wei Huangy, Jiuxing Liuz, Bulent Abaliz, Dhabaleswar K. Panday, Ohio State University, ICS ’06 Proceedings of the 20th annual international conference on Supercomputing

[v] NASA JPL Case Study, “AWS Case Study: NASA Jet Propulsion Lab’s Desert Research and Training Studies”, http://aws.amazon.com/solutions/case-studies/nasa-jpl/

[vi] Cycle Computing Case Study, “AWS Case Study: Cycle Computing and Varian”, http://aws.amazon.com/solutions/case-studies/cyclecomputing/

[vii] “Smart Computing Solutions Are On The Rise”, Forrester Research, January 2011

[viii] “HPC Applications”, Amazon.com, http://aws.amazon.com/ec2/hpc-applications/, accessed February 2011

This post appeared on Vludo’s blog a company focused on developing the platform for the next generation of product development using cloud computing and big data tools.

 

Leveraging Cloud Computing in CAE Product Development Environments (Part I)

March 23rd, 2011 No comments

Cloud computing offers significant benefits to the development of CAE engineered products including the reduction of development timelines, improvement of product performance, and significant reductions in overall development cost.  The complete benefits of cloud computing in product development environments for clients can only be realized with full participation and understanding among all stakeholders involved and careful implementation within a firms IT infrastructure.  While the full benefits of cloud based product development will only be realized with a significant transition of software and databases, partial transitions in high-performance computing and linking of interdependent disciplinary software and databases can already provide significant returns in the short term.

Client first movers in cloud based product development would likely gain a significant sustainable competitive advantage over competitors in leading technology development in this area.  The following enumerates potential benefits for both engineering and management from the transition to cloud based product development, along with potential risks and limitations.

Client Engineering Benefits

Reduced development timelines

  • Cloud based high-performance computing provides access to a virtually unlimited amount of computing power immediately as needed, eliminating computational bottlenecks
  • Dynamic scaling of computing resources provides the ability to always use appropriate compute resources for all computational workloads
  • Integrating data and software across functional silos reduces the number of iterations needed to converge on a final design

Improved product performance

  • Cloud based product development allows for a broad design of experiments on the same timeline as a single iteration, significantly increasing the probability of improved and optimal product design
  • Linking disciplinary processes provides opportunities for improved understanding and communication across functional silos resulting better optimization of products through capturing interdependencies and tradeoffs in product design
  • Shared dynamic design databases and linkages across functional software data can eliminate commonly occurring errors and wasted effort in working with non-current or incorrect data

Client Management Benefits

Cost Reductions

  • Cloud based product development reduces time spent on repeated sequential iterations to achieve design convergence and allows engineers focus on higher value added activities
  • Capital expenditures for grid computing can be reduced and converted to flexible hourly operating costs matching cyclical development cycles
  • Pay-per-use licensing models[1] allow for more direct association between costs and actual value created by product development software

Visibility and Control

  • Increased oversight, transparency, visibility, and control of the product development process through common platform usage across disciplinary teams
  • Increased ability to measure progress with clear metrics and manage local and global product development teams

[1] Leading computer aided engineering (CAE) software vendors indicate these are not only viable, but likely as product development software moves into the cloud

This post appeared on Vludo’s blog a company focused on developing the platform for the next generation of product development using cloud computing and big data tools.