The continued growth of the computational capability of throughput processors has made throughput processors the platform of choice for a wide variety of high performance computing applications. Graphics Processing Units (GPUs) are a prime example of throughput processors that can deliver high performance for applications ranging from typical graphics applications to general-purpose data parallel (GPGPU) applications. However, this success has been accompa- nied by new performance bottlenecks throughout the memory hierarchy of GPU-based systems. This dissertation identifies and eliminates performance bottlenecks caused by major sources of interference throughout the memory hierarchy. Specifically, we provide an in-depth analysis of inter- and intra-application as well as inter- address-space interference that significantly degrade the performance and efficiency of GPU-based systems. To minimize such interference, we introduce changes to the memory hierarchy for systems with GPUs that allow the memory hierarchy to be aware of both CPU and GPU applications’ charac- teristics. We introduce mechanisms to dynamically analyze different applications’ characteristics and propose four major changes throughout the memory hierarchy. First, we introduce Memory Divergence Correction (MeDiC), a cache management mecha- nism that mitigates intra-application interference in GPGPU applications by allowing the shared L2 cache and the memory controller to be aware of the GPU’s warp-level memory divergence characteristics. MeDiC uses this warp-level memory divergence information to give more cache space and more memory bandwidth to warps that benefit most from utilizing such resources. Our evaluations show that MeDiC significantly outperforms multiple state-of-the-art caching policies proposed for GPUs. Second, we introduce the Staged Memory Scheduler (SMS), an application-aware CPU-GPU memory request scheduler that mitigates inter-application interference in heterogeneous CPU-GPU systems. SMS creates a fundamentally new approach to memory controller design that decouples the memory controller into three significantly simpler structures, each of which has a separate task, These structures operate together to greatly improve both system performance and fairness. Our three-stage memory controller first groups requests based on row-buffer locality. This grouping allows the second stage to focus on inter-application scheduling decisions. These two stages en- force high-level policies regarding performance and fairness. As a result, the last stage is simple logic that deals only with the low-level DRAM commands and timing. SMS is also configurable: it allows the system software to trade off between the quality of service provided to the CPU versus GPU applications. Our evaluations show that SMS not only reduces inter-application interference caused by the GPU, thereby improving heterogeneous system performance, but also provides better scalability and power efficiency compared to multiple state-of-the-art memory schedulers. Third, we redesign the GPU memory management unit to efficiently handle new problems caused by the massive address translation parallelism present in GPU computation units in multi- GPU-application environments. Running multiple GPGPU applications concurrently induces significant inter-core thrashing on the shared address translation/protection units; e.g., the shared Translation Lookaside Buffer (TLB), a new phenomenon that we call inter-address-space interference. To reduce this interference, we introduce Multi Address Space Concurrent Kernels (MASK). MASK introduces TLB-awareness throughout the GPU memory hierarchy and introduces TLBand cache-bypassing techniques to increase the effectiveness of a shared TLB. Finally, we introduce Mosaic, a hardware-software cooperative technique that further increases the effectiveness of TLB by modifying the memory allocation policy in the system software. Mosaic introduces a high-throughput method to support large pages in multi-GPU-application environments. The key idea is to ensure memory allocation preserve address space contiguity to allow pages to be coalesced without any data movements. Our evaluations show that the MASK-Mosaic combination provides a simple mechanism that eliminates the performance overhead of address translation in GPUs without significant changes to GPU hardware, thereby greatly improving GPU system performance. The key conclusion of this dissertation is that a combination of GPU-aware cache and memory management techniques can effectively mitigate the memory interference on current and future GPU-based systems as well as other types of throughput processors.

No description available.

<p> Forest certification plays an important role in the forest products industry in Mississippi. Approximately 17% of the state&rsquo;s 19 million acres of forest land is certified under one of three major systems in the United States. More than two million acres are certified under Sustainable Forestry Initiative, more than one million acres are certified under American Tree Farm System and over 150,000 acres are certified under Forest Stewardship Council. </p><p> The goal of forest certification is sustainable use of all forest resources, from timber to clean water to recreation. This goal can only be achieved if nonindustrial private forest landowners, loggers, and forestry professionals all agree on the concepts each system espouses. </p><p> This project surveyed nonindustrial private landowners in Mississippi, members of the Mississippi Loggers Association, and members of the Mississippi Sustainable Forestry Initiative Implementation Committee, to test their levels of agreement on 12 different Likert Scales or sets of statements representing specific underlying concepts of forest certification. </p><p> The three groups generally agreed on most aspects of forest certification. There were some significant differences between the groups based on the distribution of responses, however the scale averages never differed by more than 0.7 indicating that while the differences were significant, overall they were still fairly close in their understanding of certification concepts. </p><p> There were a large number of non-industrial private landowners who were not aware of forest certification. This result has not changed since the last landowner study that was conducted in Mississippi regarding forest certification, almost 10 years ago.</p>

This thesis critically examines the concept of strategic human resource management (HRM). Existing 'critical' approaches identify prescriptive HRNI literature as 'rhetoric' that does not match 'reality'. Such an approach understands management initiatives as separate, individual, discrete, and ad hoc. However, this thesis develops an alternative perspcctivc, informed by a Foucauldian approach to 'discourse', that understands 'HRM' and 'strategy' as cultural constructs that are used by actors as they 'make sense' of discursively constructed organizational 'realities'. This perspective is then utilized to demonstrate that the existing 'critical' approaches are engaged in particular practices that define strategic HRM in a way that constructs the 'gap' between 'rhetoric' and 'reality' as HRM is simultaneously created as an academic subject. This alternative perspective provides a means of understanding and analysing prescriptive management literature and texts generated from interviews with managers in terms of two different discursively constructed 'rationalities'. Both 'rationalities' establish causal relationships between concepts of environment, organization and individual as organizational 'reality' is constructed. This perspective is utilized in the identification of the connections that are established between the managerial initiatives that are thought of as separate, individual, discrete, and ad hoc by the 'critical' literature. There are two parts to this thesis. The first part describes the development of HM4, outlines a Foucauldian conceptualization of 'discourse, and re-examines prescriptive and 'critical' HRM literature. The second part analyses texts generated from interviews with HRIpersonnel managers in a range of public and private sector organizations. This analysis demonstrates that, while there is great variety in the descriptions of organizational 'reality', connections between concepts of environment, organization and individual arc established as two key 'rationalities' are discursively constructed. It is argued that these 'rationalities' position people and practices within organizational 'reality'.

<p> An aquaculture production system was constructed at the Waik&imacr;k&imacr; Aquarium in the fall of 2016 that produces mysid shrimp zooplankton (<i> Americamysis bahia</i>). From November 2017 to February 2018, the Waik?k? Aquarium (WAq) required an average of 5,751 live mysids per week that were imported as feed with current import rates of $0.08 per animal or$24,000 annually. Fixed and variable budgets were used to evaluate the economics for the mysid culture system. Sensitivity analyses of labor costs, the sale of excess mysid production, and the discount rate were also conducted using stochastic modeling of in-house domestic yields to estimate the expected the net present values (NPV) of domestic production in comparison with imports. The indirect benefits of domestic production were qualitatively evaluated. Results showed that WAq had a greater (> 0) NPV across a variety of cost and benefit scenarios and a less expensive $.05 mysid versus the imported $.08 mysid.</p><p>

<p> Estimating the water budget of three mid-latitude extratropical cyclones is attempted from the perspective of the precipitation efficiency (PE), using a method proposed for the study of convective columns. Using a lagrangian, system-relative volume centered on the surface low pressure, each cyclone was followed for most of its lifetime within a pre-defined volume (7 degrees latitude x 9 degrees longitude, or approximately 700 x 700 km). A comparison is then made of total atmospheric water vapor ingested to total moisture eliminated (as precipitation). We hypothesize that the PE increases with the intensity of the cyclone. This small sample confirms that idea, and thus encourages further study with this approach.</p><p>

<p> Bighead Carp <i>Hypophthalmichthys nobilis</i> and Silver Carp <i> H. molitrix</i> (collectively referred to as bigheaded carp) were introduced to the United States in the 1970s and escaped into the Mississippi River from aquaculture ponds. Since their escape, bigheaded carp have become established in the Mississippi River Basin, including the Ohio River and its tributaries. More recently, bigheaded carp have invaded the Tennessee and Cumberland rivers. Little research has been conducted on bigheaded carp in Tennessee waters, with preliminary studies investigating the distribution and characteristics of established bigheaded carp populations, including growth rates and recruitment. My study represents the first systematic sampling of bigheaded carp in the Tennessee and Cumberland rivers, both tributaries to the Ohio River, and in large reservoirs in the southeast U. S. Standardized, multi-gear sampling methods were used to sample bigheaded carp in Kentucky Lake and Pickwick Lake on the Tennessee River and Lake Barkley and Cheatham Lake on the Cumberland River in 2017 and 2018. Overnight gill nets were the most effective gear, capturing an average of 17 Silver Carp per gill net gang. Short-set gill nets captured an average of one Silver Carp per gill net gang, and electrofishing captured an average of one Silver Carp per 10-minute electrofishing transect. Maximum total length (TL) was 1,390 mm for Bighead Carp and 1,111 mm for Silver Carp, and the maximum age for both species was 11 years. Average age and condition of Silver Carp was higher in upstream impoundments than downstream, and thus suggested an early invasion period from Ohio River movements. Gonadosomatic index (GSI; egg mass weight to body weight ratio) was also higher in upstream impoundments than downstream impoundments. Higher GSI values were observed in the spring and summer, when bigheaded carp may begin to spawn in response to high flows. Length-at-age estimates for the four reservoirs suggested bigheaded carp growth rates in the Tennessee and Cumberland rivers are similar to other populations in the US, indicating populations may be experiencing density effects in southeastern impoundments. No young of year (YOY) bigheaded carp were captured during the study, but weak and missing year classes due to erratic recruitment patterns are common in bigheaded carp, and monitoring for YOY fish should continue. Mean catch-per-unit-effort (CPUE) was not significantly different among reservoirs for the three gear types, and no significant seasonal variation in mean CPUE was observed for overnight gill nets. A lack of differences in CPUE across sampling seasons suggested population densities could be assessed as agencies could fit into their schedules. Overall, the results represent the first large-scale standardized sampling effort for bigheaded carp across four southeastern states and provide integral population information for natural resource managers. Future monitoring efforts should continue to track characteristics of bigheaded carp populations in the Tennessee and Cumberland rivers and determine effective control and management strategies for populations in large impoundments. </p><p>

No description available.

<p>The purpose of this study is to investigate and review literature that examines whether job satisfaction and procedural justice have a positive relationshipwith employees organisational citizenship behaviour in a brick manufacturing industry</p>

The prevalence of digital information management in an open network has driven the need to maintain balance between anonymity, authenticity and accountability (AAA). Anonymity allows a principal to hide its identity from strangers before trust relationship is established. Authenticity ensures the correct identity is engaged in the transaction even though it is hidden. Accountability uncovers the hidden identity when misbehavior of the principal is detected. The objective of this research is to develop an AAA management framework for secure resource allocations. Most existing resource management schemes are designed to manage one or two of the AAA attributes. How to provide high strength protection to all attributes is an extremely challenging undertaking. Our study shows that the electronic cash (E-cash) paradigm provides some important knowledge bases for this purpose. Based on Chaum-Pederson’s general transferable E-cash model, we propose a timed-zero-knowledge proof (TZKP) protocol, which greatly reduces storage spaces and communication overheads for resource transfers, without compromising anonymity and accountability. Based on Eng-Okamoto’s general divisible E-cash model, we propose a hypercube-based divisibility framework, which provides a sophisticated and flexible way to partition a chunk of resources, with different trade-offs in anonymity protection and computational costs, when it is integrated with different sub-cube allocation schemes. Based on the E-cash based resource management framework, we propose a privacy preserving service oriented architecture (SOA), which allows the service providers and consumers to exchange services without leaking their sensitive data. Simulation results show that the secure resource management framework is highly practical for missioncritical applications in large scale distributed information systems.