Abstract: "Directory schemes have been proposed to solve the cache coherence problem for large-scale multiprocessor systems. Most of previous studies concentrated on cost reduction for the design of directory schemes. With scalable directory design, there are various design parameters that affect its performance. Their impact is impossible to predict. In this paper, we evaluate the effect of these parameters on the performance of directory schemes concentrating on shared data, including cache organization, directory protocols, scalability and memory latency. We also analyze the resource contention and coherence delays of directory schemes and discuss possible improvements."

Many modern computer systems, including homogeneous and heterogeneous architectures, support shared memory in hardware. In a shared memory system, each of the processor cores may read and write to a single shared address space. For a shared memory machine, the memory consistency model defines the architecturally visible behavior of its memory system. Consistency definitions provide rules about loads and stores (or memory reads and writes) and how they act upon memory. As part of supporting a memory consistency model, many machines also provide cache coherence protocols that ensure that multiple cached copies of data are kept up-to-date. The goal of this primer is to provide readers with a basic understanding of consistency and coherence. This understanding includes both the issues that must be solved as well as a variety of solutions. We present both high-level concepts as well as specific, concrete examples from real-world systems. This second edition reflects a decade of advancements since the first edition and includes, among other more modest changes, two new chapters: one on consistency and coherence for non-CPU accelerators (with a focus on GPUs) and one that points to formal work and tools on consistency and coherence.

Mathematics of Computing -- Parallelism.

Today's systems are designed with Multi Core Architecture. The idea behind this is to achieve high system throuput. Once the Processor clock speed reached its saturation, designers opted for having multiple cores. Each Core or Processor equipped with their own private cache memory. But under Chip Multiprocessor, where all the processor have access to shared memory, having respective cache memory will result with Cache Coherency Problem. In Directory Protocol, for each block of data there is a directory entry that contains a number of pointers. The purpose of this number is to mention the locations of block copies. The important advantage of directory based protocols is that they scale much better than snoopy protocols. In addition to this it has the advantage of ability to exploit arbitrary point-to-point interconnects. But mean time it also has the overhead in terms of the storage and manipulation of directory state. This paper discus different Directory Based implementation, operations along with and its implementation difficulties.

An authoritative book for hardware and software designers. Caches are by far the simplest and most effective mechanism for improving computer performance. This innovative book exposes the characteristics of performance-optimal single and multi-level cache hierarchies by approaching the cache design process through the novel perspective of minimizing execution times. It presents useful data on the relative performance of a wide spectrum of machines and offers empirical and analytical evaluations of the underlying phenomena. This book will help computer professionals appreciate the impact of caches and enable designers to maximize performance given particular implementation constraints.