指导cache静态划分的程序性能profiling优化技术

对于共享cache的多核处理器,如何管理好各个核对cache的利用,对于充分发挥多核处理器性能是很关键的问题.目前采用的cache替换方法程序间会出现性能干扰,cache静态划分技术则是通过为同时运行的程序分配不同的空间来解决性能干扰问题.为了给程序分配合适大小的cache空间,需要对程序进行性能profiling,即事先多遍运行收集程序在各种cache容量下的性能数据,这种性能profiling方法开销巨大,影响实用.为了解决性能profiling需要多遍运行程序的问题,提出了只需单遍运行的程序性能profiling优化技术.该技术利用在线的phase分析技术识别程序的运行阶段,避免对相同阶段的重复profiling;同时分析程序各phase的性能同cache容量变化的关系趋势,对于性能不敏感的容量变化则不进行profiling,降低开销.在程序运行结束后通过程序各phase在cache各种容量下的性能来估计程序在各容量下的整体性能,以指导cache静态划分.实验表明,该技术的开销仅为7%,而该方法指导的cache划分比未划分时有8%的性能改进,同多遍运行的程序性能profiling指导的cache划分性能相比仅有1%的下降.     

完全图的点可区别强全染色算法

根据图的点可区别全染色的定义,结合完全图的对称性,提出一种新的点可区别强全染色算法。该算法将需要填充的颜色分为超色数和正常色数2个部分,在得到染色数量和染色次数的前提下,对超色数进行染色以增强算法收敛性。实验结果表明,该算法具有较低的时间复杂度。     

若干多重Mycielski图的邻点可区别Ⅰ-全色数

根据路和星、圈的多重Mycielski图的结构性质,用穷染递推的方法,讨论了图Mn(Cm)和Mn(Pm),以及Mn(Sm)的邻点可区别I-全染色,得到了图Mn(Sm)和Mn(Pm)的邻点可区别I-全色数等于它们的最大度,图Mn(Cm)的邻点可区别I-全色数在m=4,5时等于它的最大度加1,其余情况等于它的最大度,即分别给出图Mn(Sm)和Mn(Cm)、Mn(Pm)一种染色方案。     

DirectX在有限元分析结果三维彩色云图生成方法中的应用

本文采用扫描线法,将DirectX技术应用到有限元分析结果三维彩色云图的生成方法中,利用DirectX的实体填充模式和高洛德着色模式完成云图渲染绘制。通过实例验证,本文给出的云图生成方法不仅效果好、速度快,而且方法简单、实用。该方法适用于各类场值三维彩色云图的生成,具有很好的通用性和扩展性。     

Register allocation for write activity minimization on non-volatile main memory for embedded systems

Non-volatile memories are good candidates for DRAM replacement as main memory in embedded systems and they have many desirable characteristics. Nevertheless, the disadvantages of non-volatile memory co-exist with its advantages. First, the lifetime of some of the non-volatile memories is limited by the number of erase operations. Second, read and write operations have asymmetric speed or power consumption in non-volatile memory. This paper focuses on the embedded systems using non-volatile memory as main memory. We propose register allocation technique with re-computation to reduce the number of store instructions. When non-volatile memory is applied as the main memory, reducing store instructions will reduce write activities on non-volatile memory. To re-compute the spills effectively during register allocation, a novel potential spill selection strategy is proposed. During this process, live range splitting is utilized to split certain long live ranges such that they are more likely to be assigned into registers. In addition, techniques for re-computation overhead reduction is proposed on systems with multiple functional units. With the proposed approach, the lifetime of non-volatile memory is extended accordingly. The experimental results demonstrate that the proposed technique can efficiently reduce the number of store instructions on systems with non-volatile memory by 33% on average.     

陶瓷上仿古电镀

陶瓷上仿古电镀是将青白色的陶坯经电镀、着色,最终使之成为古色古香的陶瓷精品的工艺。本文介绍了陶瓷性能,镀前处理,有关电镀及氧化着色的工艺流程和操作规范,并详述了相应的工艺配方。     

大规模MIMO下联合用户位置和空间距离的导频分配方案

导频污染是大规模MIMO（multiple-input and multiple-output）系统的瓶颈问题,提出一种联合用户位置和空间距离的导频分配方案来抑制导频污染。该方案首先将所有用户按到达角是否重叠进行分类,将到达角互不重叠的用户归为前◢n◣类集合,剩余用户归为◢n◣+1类集合。为前◢n◣类用户集分配◢n◣组正交导频在集合内复用,而对◢n◣+1类集合中的用户进行导频选择时,定义一个干扰函数通过构建加权图的方式进行分配。同时,推导了上行频谱效率并分析了导频开销以及算法复杂度。仿真表明,该方案不仅大大减少了算法复杂度和导频开销,而且显著提高了系统的频谱效率（spectral efficiency,SE）。     

A novel nest-based scheduling method for mobile wireless body area networks

Wireless Body Area Networks (WBANs) comprise various sensors to monitor and collect various vital signals, such as blood pressure, pulse, heartbeat, body temperature, and blood sugar. A dense and mobile WBAN often suffers from interference, which causes serious problems, such as wasting energy and degrading throughput. In reality, not all of the sensors in WBAN need to be active at the same time. Therefore, they can be divided into different groups so that each group works in turn to avoid interference. In this paper, a Nest-Based WBAN Scheduling (NBWS) algorithm is proposed to cluster sensors of the same types in a single or multiple WBANs into different groups to avoid interference. Particularly, we borrow the graph coloring theory to schedule all groups to work using a Time Division for Multimodal Sensor (TDMS) group scheduling model. Both theoretical analysis and experimental results demonstrate that the proposed NBWS algorithm performs better in terms of frequency of collisions, transmission delay, system throughput, and energy consumption compared to the counterpart methods.     

一类国产复杂异构系统上的HPCG并行算法及高效实现研究

HPCG基准测试程序是一种新的超级计算机排名度量标准.该测试基准主要用于衡量超级计算机解决大规模稀疏线性系统的能力,更贴近实际应用,近年来广受关注.基于国产超级计算机研究异构众核并行HPCG软件具有非常重要的意义,其不仅可以提升国产超级计算机HPCG的排名,还对很多应用提供了并行算法、优化技术等方面的参考.本文面向某国产复杂异构超级计算机开展研究,首先采用了分块图着色算法对HPCG进行并行,并提出一种适用于结构化网格的图着色算法,该算法并行性能高于传统的JPL、CC等算法,且着色质量高,运用于HPCG后,迭代次数减少了3次,整体性能提升了6%.本文还分析了复杂异构系统各个部件传输的开销,提出一套更适用于HPCG的任务划分方法,并从稀疏矩阵存储格式、稀疏矩阵重排、访存等角度开展了细粒度的优化.另外在多进程计算时,还采用了内外区划分算法将核心函数SpMV、SymGS中的邻居通信操作进行了隐藏.最终整机测试时,性能达到国产超级计算机峰值性能的1.67%,相比单节点,整机弱可扩展性并行效率达到了92%.     

大规模MIMO系统中基于图着色的半动态导频分配

针对大规模MIMO系统中存在的导频污染问题,提出了一种基于图着色的半动态导频分配方案。首先,推导出各个小区可以复用导频的用户与其服务基站的最大距离门限。然后,根据该门限值将系统内各小区的用户分为小区中心用户和小区边缘用户;允许各个小区门限值之内的用户（小区中心用户）复用相同的导频;最后,基于图着色理论协同分配各个小区的小区边缘用户的导频,达到降低导频干扰的目的。仿真结果显示,该方案减少了用户间的导频干扰,有效地提升了整个系统的上行可达和速率。