We present a unified approach to locality optimization that employs both data and control transformations. Data transformations include changing the array layout in memory. Control transformations involve changing the execution order of programs. We have developed new techniques for compiler optimizations for distributed shared-memory machines, although the same techniques can be used for sequential machines with a memory hierarchy. Our compiler optimizations are based on an algebraic representation of data mappings and a new data locality model. We present a pure data transformation algorithm and an algorithm unifying data and control transformations. While there has been much work on control transformations, the opportunities for data transformations have been largely rleglected. In fact, data transformations have the advantage of being applicable to programs that cannot be optimized wiith control transformations. The unified algorithm, which performs data and control transformations simultaneously, offers improvement over optimizations obtained by applying data and control transformations separately. The experimental results using a set of applications on a parallel machine show that the new optimizations improve performance significantly. These results are further analyzed using locality metrics with instrumentation and simulation.

No abstract

Ovarian cancer ranks fifth in cancer deaths among women, accounting for more deaths than any other cancer of the female reproductive system. Monocyte chemoattractant protein-1 (MCP-1) is highly expressed in various malignancies and promotes carcinogenesis. The aim of the study was to investigate the association between MCP-1 genetic polymorphisms and the susceptibility to ovarian cancer. MCP-1 rs1024611A/G and rs3760396C/G polymorphisms were examined in 257 ovarian cancer patients and 273 healthy controls. We found that distributions of rs1024611GG genotype and rs3760396GG genotype were clearly increased in ovarian cancer cases compared to healthy donors (odds ratio [OR]=1.93, 95% confidence interval [CI]: 1.13-3.29, p=0.015; OR=3.89, 95% CI: 1.63-9.33, p=0.001). Stratification analyses revealed that patients with serous papillary type had further increased percentage of rs3760396GG genotype than those with other types (OR=3.89, 95% CI: 1.11-13.66, p=0.024). In addition, we evaluated the possible effect of MCP-1 polymorphisms on gene expression by examining the serum level of MCP-1 in patients and controls. Data revealed that subjects carrying rs1024611AG and GG genotypes had a significantly higher serum level of MCP-1 than those with AA genotype. These data suggest that MCP-1 rs1024611A/G and rs3760396C/G polymorphisms are associated with increased susceptibility to ovarian cancer, in which rs1024611A/G may increase serum level of MCP-1 in the Chinese population.

The Greater Toronto Area is the most vital economic centre in Canada and has experienced rapid urban expansion in the past 40 years. This research uses Landsat images to detect the spatial and temporal dynamics of urban expansion in the Greater Toronto Area from 1974 to 2014. We quantitatively analyzed the extent of urban expansion and spatial patterns of growth from classified Landsat images. We then integrated our expansion findings with population data to observe the relationships between urban growth and population. We found that the Greater Toronto Area had significant growth of 1115 km2, expanding mainly in radiated and ribbon expansion modes. There was substantial correlation between urban extent and population in the period of study. These results demonstrate the efficacy of combining statistical population data with remote sensing imagery for the analysis of urban expansion.

An experimental investigation was conducted on the single-tooth pulsating bending fatigue of 38SiMnMo alloy steel at a frequency range of 140-150 Hz. Using both up-down test method and S-N curve extension method, statistical data in terms of curves and endurance limits were analysed. The testing and analytical results indicated that the 'platform' of fatigue curves exists and the S-N curve cannot be extended to the cyclic basic number N b in order to obtain the endurance limit. Furthermore, on the basis of the life distribution optimization, the three-parameter Weibull distribution was used for statistical data analysis. The statistical analysis showed that the bending endurance limit of 38SiMnMo alloy steel gears estimated at 99 per cent reliability and 95 per cent confidence was 284 MPa, which was located at the slightly low part of the mean line in the chart range illustrated in the standards ISO6336 and GB3480. The results demonstrated that the method described in this article was valuable.