The Experts below are selected from a list of 38901 Experts worldwide ranked by ideXlab platform
Jianlong Zhong - One of the best experts on this subject based on the ideXlab platform.
-
kernelet high throughput gpu kernel executions with dynamic slicing and scheduling
IEEE Transactions on Parallel and Distributed Systems, 2014Co-Authors: Jianlong ZhongAbstract:Graphics processors, or GPUs, have recently been widely used as accelerators in shared environments such as clusters and clouds. In such shared environments, many kernels are submitted to GPUs from different users, and throughput is an important metric for performance and total Ownership Cost. Despite recently improved runtime support for concurrent GPU kernel executions, the GPU can be severely underutilized, resulting in suboptimal throughput. In this paper, we propose Kernelet, a runtime system to improve the throughput of concurrent kernel executions on the GPU. Kernelet embraces transparent memory management and PCI-e data transfer techniques, and dynamic slicing and scheduling techniques for kernel executions. With slicing, Kernelet divides a GPU kernel into multiple sub-kernels (namely slices ). Each slice has tunable occupancy to allow co-scheduling with other slices for high GPU utilization. We develop a novel Markov chain-based performance model to guide the scheduling decision. Our experimental results demonstrate up to 31 percent and 23 percent performance improvement on NVIDIA Tesla C2050 and GTX680 GPUs, respectively.
-
kernelet high throughput gpu kernel executions with dynamic slicing and scheduling
arXiv: Distributed Parallel and Cluster Computing, 2013Co-Authors: Jianlong ZhongAbstract:Graphics processors, or GPUs, have recently been widely used as accelerators in the shared environments such as clusters and clouds. In such shared environments, many kernels are submitted to GPUs from different users, and throughput is an important metric for performance and total Ownership Cost. Despite the recently improved runtime support for concurrent GPU kernel executions, the GPU can be severely underutilized, resulting in suboptimal throughput. In this paper, we propose Kernelet, a runtime system with dynamic slicing and scheduling techniques to improve the throughput of concurrent kernel executions on the GPU. With slicing, Kernelet divides a GPU kernel into multiple sub-kernels (namely slices). Each slice has tunable occupancy to allow co-scheduling with other slices and to fully utilize the GPU resources. We develop a novel and effective Markov chain based performance model to guide the scheduling decision. Our experimental results demonstrate up to 31.1% and 23.4% performance improvement on NVIDIA Tesla C2050 and GTX680 GPUs, respectively.
Tareq Z Ahram - One of the best experts on this subject based on the ideXlab platform.
-
Economics of human performance and systems total Ownership Cost
Work, 2012Co-Authors: Wilawan Onkham, Waldemar Karwowski, Tareq Z AhramAbstract:Financial Costs of investing in people is associated with training, acquisition, recruiting, and resolving human errors have a significant impact on increased total Ownership Costs. These Costs can also affect the exaggerate budgets and delayed schedules. The study of human performance economical assessment in the system acquisition process enhances the visibility of hidden Cost drivers which support program management informed decisions. This paper presents the literature review of human total Ownership Cost (HTOC) and Cost impacts on overall system performance. Economic value assessment models such as Cost benefit analysis, risk-Cost tradeoff analysis, expected value of utility function analysis (EV), growth readiness matrix, multi-attribute utility technique, and multi-regressions model were introduced to reflect the HTOC and human performance-technology tradeoffs in terms of the dollar value. The human total Ownership regression model introduces to address the influencing human performance Cost component measurement. Results from this study will increase understanding of relevant Cost drivers in the system acquisition process over the long term.
-
Human Total Ownership Cost in Complex Systems Estimating Human Performance Cost in Environments with Multiple Layers of Technology
Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 2011Co-Authors: Tareq Z Ahram, Waldemar Karwowski, Wilawan OnkhamAbstract:In current practice, the Human Total Ownership Cost (HTOC) is often viewed as an acquisition Cost—that is, the Cost of systems that does not consider human Costs other than those related to the immediate staffing, training needs and requirements. The Cost of human performance, in terms of both human capabilities and limitations, is directly related to the total Ownership Cost (TOC) of the technology. This paper summarizes ongoing efforts for modeling complex human performance Cost drivers, for identification of critical elements that define the long term impact of human performance on system design, development, production, fielding, sustainment, and improvement, throughout the system life cycle. The overall aim is to develop a methodology to better assess and predict the total Cost of human and human performance as relevant to acquisition of new technologies and systems. Future challenges that decision makers are likely to face in this domain are also discussed.
Brad R. Naegle - One of the best experts on this subject based on the ideXlab platform.
-
Total Ownership Cost a Decade Into the 21st Century
2012Co-Authors: Brad R. Naegle, Michael W. BoudreauAbstract:Abstract : The intent of this research is to gather together the various approaches for controlling and reducing total Ownership Cost (TOC), and to describe tools and methods to assist Program Managers and others in addressing TOC more effectively. This study examines TOC from the perspective of congressional direction, the perspective of the Office of the Secretary of Defense and Service leadership's governance, the perspective of Program Manager execution, and the perspective of available infrastructure support.
-
Total Ownership Cost Considerations in Key Performance Parameters and Beyond (Defense Acquisition Review Journal)
2005Co-Authors: Michael W. Boudreau, Brad R. NaegleAbstract:Abstract : Total Ownership Cost (TOC) is an initiative to manage Costs over the entire life cycle of a system, and the Program Manager (PM) is responsible for estimating and managing system TOC using input from key stakeholders. There are several major categories of Costs that contribute to TOC. For instance, the phases of a weapon system's life cycle may include research and development, production, operating and support, and disposal. This paper addresses ways to encourage a Life Cycle Cost (LCC) perspective, examine the critical issues associated with understanding and implementing the TOC concept, and assist the PM with the knowledgeable execution of a TOC plan. Also, metrics necessary to ensure appropriate implementation are explored, various methods of controlling and reducing TOC are evaluated, and areas where the Department of Defense needs to refocus are addressed.
-
Total Ownership Cost considerations in key performance parameters and beyond
2005Co-Authors: Michael W. Boudreau, Brad R. NaegleAbstract:Total Ownership Cpst (TOC) is an initiative to manage Costs over the entire life cycle of a system, and the Program Manager (PM) is responsible for estimating and managing system TOC using input from key stakeholders. There are several major categories of Costs that contribute to TOC. For instance, the phases of a weapon system's life cycle may include research and development, production, operating and support, and disposal. This paper addresses ways to encourage a Life Cycle Cost (LCC) perspective, examine the critical issues associated with understanding and implmenting the TOC concept, and assist the PM with the knowledgeable execution of a TOC plan. Also metrics necessary to ensure appropriate implementation are explored, various methods of controlling and reducing TOC are evaluated, and areas where the Department of Defense needs to refocus are addressed.
-
Total Ownership Cost: An Exercise in Discipline
2004Co-Authors: Michael W. Boudreau, Brad R. NaegleAbstract:Abstract : This paper presents number of recommendations regarding Total Ownership Costs (TOC): (1) Designate the TOC target as a System Key Performance Parameter (KPP) -- causes the Program Manager to focus on and continually report on TOC, places TOC on equal footing with performance, facilitates trade-off analysis in favor of TOC, and focuses contractors on TOC-efficient design; (2) Establish TOC databases for tracking TOC Cost drivers and measuring R-TOC effectiveness; (3) Establish contractor incentives for improving system TOC characteristics; and (4) Garner user/upper leadership support for TOC goals.
-
The Impact of Software Support on System Total Ownership Cost
2004Co-Authors: Brad R. NaegleAbstract:Abstract : The software component of our increasingly high-technology weapon systems provides the capabilities and lethality desired for our forces, but it is potentially devastating to our ability to Cost-effectively maintain their advantages. The complexity of individual software-intensive systems is significantly compounded when they are combined in a "system of systems" architecture. The initial software architecture, driven by how requirements are translated into performance specifications, is critical in determining how much maintenance will be required and how much effort will be required in the necessary maintenance actions. To gain more effective software design, significantly more effort is required in requirements analyses. Performance specifications must be much more developed than is typical in the current development model. Handing off performance specifications developed through just three levels of the Work Breakdown Structure (WBS) for software-intensive systems is insufficient in a complex, system of systems environment dependent on seamless interfaces in an ever-changing architecture. Significant development, incorporating all critical performance features, interface requirements, and known, planned, and projected upgrades, changes, and enhancements must be effectively transmitted to the developer for consideration in the software design and architecture. Without these efforts, software supportability Costs will continue to skyrocket as existing software will require expensive and time-consuming re-engineering to accommodate interface and capability changes that were known or could have been derived from a more thorough requirements analyses.
Wilawan Onkham - One of the best experts on this subject based on the ideXlab platform.
-
Economics of human performance and systems total Ownership Cost
Work, 2012Co-Authors: Wilawan Onkham, Waldemar Karwowski, Tareq Z AhramAbstract:Financial Costs of investing in people is associated with training, acquisition, recruiting, and resolving human errors have a significant impact on increased total Ownership Costs. These Costs can also affect the exaggerate budgets and delayed schedules. The study of human performance economical assessment in the system acquisition process enhances the visibility of hidden Cost drivers which support program management informed decisions. This paper presents the literature review of human total Ownership Cost (HTOC) and Cost impacts on overall system performance. Economic value assessment models such as Cost benefit analysis, risk-Cost tradeoff analysis, expected value of utility function analysis (EV), growth readiness matrix, multi-attribute utility technique, and multi-regressions model were introduced to reflect the HTOC and human performance-technology tradeoffs in terms of the dollar value. The human total Ownership regression model introduces to address the influencing human performance Cost component measurement. Results from this study will increase understanding of relevant Cost drivers in the system acquisition process over the long term.
-
Human Total Ownership Cost in Complex Systems Estimating Human Performance Cost in Environments with Multiple Layers of Technology
Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 2011Co-Authors: Tareq Z Ahram, Waldemar Karwowski, Wilawan OnkhamAbstract:In current practice, the Human Total Ownership Cost (HTOC) is often viewed as an acquisition Cost—that is, the Cost of systems that does not consider human Costs other than those related to the immediate staffing, training needs and requirements. The Cost of human performance, in terms of both human capabilities and limitations, is directly related to the total Ownership Cost (TOC) of the technology. This paper summarizes ongoing efforts for modeling complex human performance Cost drivers, for identification of critical elements that define the long term impact of human performance on system design, development, production, fielding, sustainment, and improvement, throughout the system life cycle. The overall aim is to develop a methodology to better assess and predict the total Cost of human and human performance as relevant to acquisition of new technologies and systems. Future challenges that decision makers are likely to face in this domain are also discussed.
D.c. Frost - One of the best experts on this subject based on the ideXlab platform.
-
Study of an amorphous alloy core transformer
Journal of Magnetism and Magnetic Materials, 1994Co-Authors: Andrzej Nafalski, D.c. FrostAbstract:Abstract Amorphous core transformers (ACT) have become a technological and commercial reality and there are an estimated 400,000 units installed worldwide [1]. Their applications reflect changes in buying practices, where the efficiency evaluation is an important factor in the purchasing decision for distribution transformers. Use of the total Ownership Cost (TOC) concept facilities the selection of a transformer on the basis of its performance. This concept is used in this paper to investigate the feasibility of applying a distribution ACT in Western Australian (WA). A 10 kVA ACT, evaluated by the TOC method, was compared with a traditional silicon iron core transformer of the same rating. The Cost of amorphous metal (relative to alternative materials), the distribution load profile, and the values of capitalised loss Costs are factors which affect the Cost effectiveness of ACTs.
