存储层次

GPU异构系统中的存储层次和负载均衡策略研究

Research on Memory Hierarchy and Load Balance Strategy in Heterogeneous System Based on GPU

提出了一种以大容量流寄存器堆替代Cache作为中间缓冲器，并能适合于媒体处理应用的金字塔存储层次体系结构设计。

A new architecture design of pyramid memory hierarchy was brought forward , which substituted the larger stream register files for Cache as the medial buffers and could adapt to media processing applications .

至少在数据存储层次上，不能使用表连接（Join），也不要定义关联完整性约束。

No joins , and no referential integrity , at least at the data store level .

本文分析了Internet应用的特点，并且在不同的存储层次上给出了Internet应用下存储系统的特点。

In this paper , we carefully analyzed the properties of internet applications , and explained the requirement on storage in three different layers .

第一部分设计了一种高效能、自适应的缓存替换策略，用来管理存储层次中的DRAM。

Firstly , we design a high performance 、 adaptive cache replacement policy to administer the data in DRAM .

现代的高性能计算机体系结构中，为了更有效地实现计算与访存的平衡，优化访存带宽和延迟，越来越多地采用软件管理的多级存储层次来替代硬件管理的多级cache存储层次。

In modern high performance architecture , to balance the computation and memory access , optimize the memory bandwidth and latency , software-managed memory hierarchy is increasingly used instead of hardware-managed cache hierarchy .

动态可重构Cache的思想最早是罗彻斯特大学（UniversityofRochester）的学者在他们的一篇关于存储层次的论文1中提出的，当时主要是针对高性能的超标量通用处理器。

The idea of dynamically reconfigurable cache originally appeared in a paper about memory-hierarchy , which is written by scholar of University of Rochester 1 . And they aim at high performance superscalar general-purpose processor .

自动分析I/O访问特征，预测其后续操作，主动把可能需要的对象预取到PCM高速存储层次。

The OSD can analyze the I / O access pattern , predict the subsequent operations , and pre-fetch the possible objects into the high-speed storage layer based on PCM .

PLG方法采用本文提出的DLRS数据结构表示模型，以基网格和操作序列的形式表示及存储层次细节。

PLG method uses DLRS data structure to represent meshes , store level of detail by basic mesh and operation sequence .

本文针对多级SPM存储层次下片上多核体系结构中的并行程序优化方法进行了研究，提出了三种启发式任务与数据关联调度算法。

The paper mostly focuses on the study of parallel optimization of the on-chip multi-processor system with multi-level scratchpad memory hierarchy , and proposed three algorithms with heuristic policy for associated scheduling of task and data .

处理器速度比内存速度增加得更快导致了所谓的处理器与内存墙问题，为了弥补这种差距，现代计算机主要依赖一个具有容量小而速度快的cache的存储层次组织。

Processor speed has been increasing at a much greater rate than memory speed leading to the so called processor-memory gap . In order to compensate for this gap in performance , modern computers rely heavily on a hierarchical memory organization with a small amount of fast memory called cache .

文本方式存储层次型结构数据的方法及处理

Processing Method of Storing Hierarchical data structure in Text Mode

计算机工程师们采用存储层次来解决这个鸿沟。

Therefore the computer engineers use memory hierarchy to solve this question .

一种新型网络视频点播的存储层次

New Memory Hierarchies for Network Video - On-Demand

一种多核处理器存储层次性能评估模型

An Evaluation Model for Multi-Core Processor Memory Hierarchy

基于经验搜索的多级存储层次优化

Memory Hierarchy Optimization Based on Empirical Search

计算机存储层次模型

Computer storage hierarchy mode

本文扩展了图着色寄存器分配算法，首次将其运用到多级存储层次分配上。

This thesis extends the graph coloring algorithm , for the first time , applies it to multi-level memory allocation .

通过将存储层次建模为一个带权图，我们的方法可以运用到任何多级软件管理存储层次组织上。

By modeling the memory hierarchy as a weighted graph , our method is applicable to any software-managed memory hierarchy organization .

针对这些问题和缺陷，人们提出了软件管理存储层次方案，并进行了广泛研究。

To solve these problems , the software-managed memory hierarchy has been proposed , which has attracted many researchers ' attention .

分片式流处理器的性能是否也具有可扩展性依赖于其支持的编程模型、片上存储层次、片上互连网络以及计算模型。

The performance of tiled stream processor is determined by programming model , memory hierarchy , NoC ( Networ-on-Chip ) and computation model .

我们对传统数据相干图进行扩展，提出路径合并和路径消解技术，有效地减少存储层次间通信。

We also extend the concept of data interference graph , propose the path merging and path reduction techniques to reduce the inter-level communication .

寄存器文件作为微处理器内核的关键部件，处于存储层次的顶层，是访问速度要求最高的存储部件。

Register File is a key part in microprocessor , located in the top level of memory hierarchy , so it requires the highest access speed .

遗憾的是，由于不同数据集中关系表的模式在存储层次上往往不相同，彼此之间的冗余度很小，导致按行存储的数据存储方式并不利于数据复用的实现。

Unfortunately , due to schemas in different data sets on memory hierarchy is often not the same , the redundancy is very little . That results it is not very easy to realize row-stores ' data reusing .

根据数据访问请求处理的生产&消费特性，使用排队理论建立了一种存储层次评估模型，用于分析存储层次的配置，指导存储层次优化。

Based on the ' producer-consumer ' property of the data access request handling , a memory hierarchy model is built using the queue theory to analyze the configuration of memory hierarchy and to guide the optimization of memory hierarchy .

论文提出基于经验搜索的多级存储层次优化方法，将优化多级存储层次问题转化为对优化参数的经验搜索问题，并基于遗传算法选择全局最优解。

We propose the approach of Memory Hierarchy Optimization based on Empirical Search . It turns the problem of optimizing across multiple levels of the memory hierarchy to an empirical search to optimization parameter problem , and selects the best overall solution .

针对流应用中各种可用的并行性和局域性，我们设计并实现了由异构处理器共享的二级缓存子系统，并对片外存储层次进行改进，完善了原有的存储层次结构。

Then , we design and implement a Shared by Heterogeneity Processor L2 Cache sub-system to exploit the extra parallelism and locality in stream application , and improve the off-chip memory bandwidth . These lead to a more completed parallel hierarchy stream memory system .

存储芯片层次化分割P/G网等效电阻求解算法

A Hierarchy Partition Algorithm for Calculating the Equivalent Resistance in Memory Chip P / G Network

正如前面提到的，DB2允许根据XML列中存储的层次化数据创建关系视图。

As mentioned previously , DB2 enables firms to create relational views over hierarchical data stored in XML columns .

DB2pureXML在可同时理解XML和关系数据的混合数据库中引入了使用原生方法存储XML层次结构数据的功能。

DB2 pureXML introduces the ability to natively store XML hierarchical data within a hybrid database that understands XML as well as relational data .