This paper proposes a new architecture, called Adaptive PREfetching and Scheduling (APRES), which improves cache efficiency of GPUs. APRES relies on the observation that GPU loads tend to have either high locality or strided access patterns across warps. APRES schedules warps so that as many cache hits are generated as possible before the generation of any cache miss. Without directly predicting future cache hits/misses for each warp, APRES creates a warp group that will execute the same static load shortly and prioritizes the grouped warps. If the first executed warp in the group hits the cache, grouped warps are likely to access the same cache lines. Unless, APRES considers the load as a strided type and generates prefetch requests for the grouped warps. In addition, APRES includes a new dynamic L1 prefetch and data cache partitioning to reduce contentions between demand-fetched and prefetched lines. In our evaluation, APRES achieves 27.8 percent performance improvement.

This paper proposes a new data prefetching technique for Graphics Processing Units (GPUs) called Warp Aware Selective Prefetching (WASP). The main idea of WASP is to dynamically select warps whose progress is slower than that of the current warp as prefetching target warps. Under the in-order instruction execution model of GPUs, these prefetching target warps will certainly execute the same load as the current warp. Exploiting that, WASP prefetches the data for prefetching target warps, which allows the prefetched data to be accurately accessed. To simply verify the progress of the warps, WASP monitors the counts of the dynamic load executions for all warps. When a warp executes a load, WASP searches the warps with lower load execution counts than the current warp and generates the prefetch requests for them. In our evaluation, WASP achieves a 16.8 percent speedup compared to the baseline GPU.

Recently, interactive media which tries to maximize the audience engagement has been widely distributed throughout the digital media environment. For example, developing and promoting a new participatory of media genre with higher immersion which can be applied to advertisement, film, game, and e-learning are actively researched. In this paper, we introduce a new interactive media authoring system that enables bidirectional communication between the user and the applications. This authoring system provides a modality interface which manages interactions between various modalities, i.e. which allows users to choose one or more modalities as inputs and to get other modalities as outputs. This system helps the user to engage with the media and minimizes the image producer-user constraints in various distribution enabled web environment via PC and smart devices, without additional equipment installation.

Long memory latency and limited throughput become performance bottlenecks of GPGPU applications. The latency takes hundreds of cycles which is difficult to be hidden by simply interleaving tens of warp execution. While cache hierarchy helps to reduce memory system pressure, massive Thread-Level Parallelism (TLP) often causes excessive cache contention. This paper proposes Adaptive PREfetching and Scheduling (APRES) to improve GPU cache efficiency. APRES relies on the following observations. First, certain static load instructions tend to generate memory addresses having very high locality. Second, although loads have no locality, the access addresses still can show highly strided access pattern. Third, the locality behavior tends to be consistent regardless of warp ID. APRES schedules warps so that as many cache hits generated as possible before any cache misses generated. This is to minimize cache thrashing when many warps are contending for a cache line. However, to realize this operation, it is required to predict which warp will hit the cache in the near future. Without directly predicting future cache hit/miss for each warp, APRES creates a group of warps that will execute the same load instruction in the near future. Based on the third observation, we expect the locality behavior is consistent over all warps in the group. If the first executed warp in the group hits the cache, then the load is considered as a high locality type, and APRES prioritizes all warps in the group. Group prioritization leads to consecutive cache hits, because the grouped warps are likely to access the same cache line. If the first warp missed the cache, then the load is considered as a strided type, and APRES generates prefetch requests for the other warps in the group. After that, APRES prioritizes prefetch targeted warps so that the demand requests are merged to Miss Status Holding Register (MSHR) or prefetched lines can be accessed. On memory-intensive applications, APRES achieves 31.7% performance improvement compared to the baseline GPU and 7.2% additional speedup compared to the best combination of existing warp scheduling and prefetching methods.

In distributed query processing, good estimation algorithms of communication costs are critical for query processing, including distributed XML queries. There are techniques that estimate a communication cost for distributed SQL query processing, and some of techniques are adopted in numerous distributed SQL processors. Therefore adopting the processing techniques for SQL queries for the communication cost‐based processing of the distributed XML queries seems natural. Unfortunately, however, the tree‐structured XML document is different from the table‐shaped relational data. These structural differences make adopting the techniques for SQL queries difficult. This study defines some of the considerations for estimating the communication cost of distributed XML queries, and proposes a method for communication cost‐based query processing. The experiments show that the proposed algorithm is reasonable for estimating the communication cost for distributed XML queries.

XML has been established as a format for data exchange on the Internet, which has a large volume scale. By the increase of the volume of XML data, applications need to have a standard query language for efficient retrieving and integrating XML data. XQuery, which is proposed by W3C, is one of the standard query languages for XML data which is widely accepted by many applications. In relation to this, the efficient processing of XQuery queries has become a new research topic and a few researchers have proposed the methods for optimizing XQuery queries to resolve the problem of efficiency. A number of previous researches refer to other information like XML schema or DTD for optimizing the queries. However, our optimization technique does not use any other information except the XQuery query instance itself. Also, some researches just define system dependant optimization methods. In other words, they optimize the execution plan of XQuery queries therefore their methods depend on their XQuery execution engine. However our method can be used in any XQuery engine and in any middleware like a mediator over distributed XQuery engines, because our input is an XQuery query and output also is an optimized XQuery query- The express power of XQuery language is abundant but its syntax is very complex because XQuery borrowed features from several other languages. Therefore, we think that it is necessary to classify the queries and give an optimization method according to the characteristics of classified queries. For these reasons, we consider FOR clauses in XQuery queries because FOR clauses can generate some redundant execution of unnecessary operations. In this paper, we classify the general cases of XQuery queries including FOR clauses, and propose the rewriting rules based on these cases. The focus of our optimization is to reduce the frequency of iterations generated by FOR clauses, of course, without semantic modification of the original query. We implement a prototype system for our optimization algorithm and evaluate the performance between original XQuery queries and optimized XQuery queries.

Multimedia GIS constitutes a very promising new user interface concept for ubiquitous environments of GIS applications. Geospatial video services will be an important class of application for users of ubiquitous GIS or LBS. A novel approach is presented for linking between geospatial videos and nonspatial data using remotely sensed data. We have identified three key challenges (georeferencing; content creation for geospatial videos; bidirectional linking) that should be addressed to enable linking among geographic hypermedia such as 2D maps, 3D virtual environments and geospatial videos. Experimental results indicate that the proposed approach is effective in retrieving geospatial video clips and nonspatial data

Urban facility management is one of the applications to which the virtual GIS is well applied. For virtual GIS to be utilized to manage urban facilities, some functionalities should be provided. One of them is how to manage the huge size of 3D geographic data as compact as possible at the same time providing realistic modeling and visualization. The other one is spatial analysis on urban facilities especially on multistorey buildings. In this paper, the synthetic modeling approach is proposed to model urban facilities realistically from the compact size of GIS data. Geographic features positioned in each storey of a building are also modeled and managed hence they are identified and queried for spatial analysis.

Proposes a new terrain visualization technique associated with the conversion process from a contour-based data model to a triangulated irregular network (TIN) model. The main research related to processing these contour lines had been focused on interpolation methods to create the digital terrain model (DTM). We introduce a novel approach for visualizing terrain by using contour line data which are obtained from the digital elevation model (DEM). Our approach uses the marching square algorithm to extract contour lines from the DEM and classify contours into open contours and closed contours to calculate local minimum/maximum points. To minimize the search space for calculating local minimum/maximum, we construct the containment tree of the closed contour. Then we only achieve an approximation procedure for calculating minimum/maximum points at leaf nodes in the containment tree. Finally, we can easily reconstruct the terrain by performing constrained Delaunay triangulation that uses approximation points and contour data as inputs. In the experiments, we compare our algorithm with DEM processing in terms of the processing speed and data size.