Blue Collar Computing
From Wikipedia, the free encyclopedia
 |
This article or section is written like an advertisement. Please help rewrite this article from a neutral point of view per Wikipedia policy. Mark blatant advertising for speedy deletion with {{db-spam}}. (help, talk) |
Blue Collar Computing (BCC) is high performance computing (HPC) for industries that do not currently have the time, money or expertise to invest in HPC resources. Larger industries already know the important role that HPC plays in their success, the The Ohio Supercomputer Center (OSC) is working to bring the power of HPC to small to midsize companies.
To stay competitive in both the field of science and the world market the United States needs:
- easy to use HPC programming languages
- comprehensive HPC training
- a collaborative cyberinfrastructure
Creating and testing these types of languages, providing HPC training and building this cyberinfrastructure are the primary goals of Blue Collar Computing.
Why Blue Collar Computing? American manufacturers are under continuous pressure to deliver goods at lower price points and with improved quality. A key technology for achieving these goals is the use of computational models and simulations at both the engineering design level as well as the manufacturing level. Modeling in design processes allow engineers to explore a greater area of the possible "design space" for a product and deliver better designs in shorter times. Manufacturers can use simulations to determine the behavior of materials during the manufacturing process and, based on that knowledge, improve efficiency by adjusting their processes to improve yield and reduce waste.
Large companies have found a competitive advantage from such simulations by leveraging high-performance computing (HPC) systems and High Performance Technical Computing (HPTC) software. For example, General Motors is using parallel computing to simulate the crash testing of automobiles. Genaral Motors claims that the digital computer crash simulations, made possible due to HPC hardware and HPTC software, can reduce the number of full-size crash vehicle tests by more than 85%, at a cost of $500,000 per test.
Unfortunately, only the largest firms have realized the advantages of HPTC-based simulations. These firms had to make substantial investments to: 1. Develop in-house computational modeling and simulation expertise 2. Purchase the required HPTC software 3. Purchase an HPC system (or HPC machine time from a shared center)
These "barriers to entry" place an undue burden on small and medium sized firms. This, in turn, leads to underutilization of computational simulations that can lower costs and improve efficiency.
Blue Collar Computing (BCC) is an approach to supplying computational modeling and simulation solutions to companies without requiring large up-front investment. The BCC model is service-oriented, relying on amortizing the cost of expertise, software and hardware across many firms to reduce individual expenditures. The BCC model will enable the majority of U.S. firms to exploit numerical models and simulations.
How does Blue Collar Computing work? In the BCC model, the client is a manufacturing or engineering firm that wishes to use computational modeling and simulation. These simulations may be run at varying levels of fidelity based on the criteria under study.
In this model, the simulation has several components:
1. Model Interface - The model interface is the user interface for the simulation application. An analyst (i.e., domain expert) creates the computational model and develops the model interface using knowledge of the client’s manufacturing process.
2. Small-Scale Solver - (Optional) HPTC software used to execute some of the numerical simulations. Provides faster turnaround time for small jobs.
3. HPC Portal Client - Together with the HPC Portal Server, provides middleware services between the client site and the HPC resource center. Handles transmission of job requests/results between sites.
4. HPC Portal Server - See above.
5. HPC Job Manager - Scheduler for shared systems at HPC center.
6. Large-scale solver - HPTC software used for computational simulation. Provides same functionality as small-scale solver, but may employ algorithms that have greater parallel scalability.
