Software Support Memory Forwarding
نویسندگان
چکیده
Memory forwarding is an effective way to dynamically optimize the data layout. It provides a safe way to improve the performance of cache by actively creating better data locality. The existing memory forwarding mechanism needs additional hardware support to perform the address redirection, which greatly limits its real application. To put the memory forwarding technique into widely practice, we propose a new way of software -based implementation. Based on the fact that good data layout can bring much improvement of cache performance and actual forwarding is a rare event during the execution, our experiment shows that reasonable speedup that can be achieved by using this software-based memory forwarding technique .
منابع مشابه
Data Forwarding in Scalable Shared-Memory Multiprocessors1
Scalable shared-memory multiprocessors are often slowed down by long-latency memory accesses. One way to cope with this problem is to use data forwarding to overlap memory accesses with computation. With data forwarding, when a processor produces a datum, in addition to updating its cache, it sends a copy of the datum to the caches of the processors that the compiler identified as consumers of ...
متن کاملOn Closed Nesting in Distributed Transactional Memory
Distributed Software Transactional Memory (D-STM) is a recent but promising model for programming distributed systems. It aims to present programmers with a simple to use abstraction (transactions), while maintaining performance and scalability similar to distributed fine-grained locks. Any complications usually associated with such locks (i.e. distributed deadlock) are avoided. Building upon t...
متن کاملA High Performance Parallel IP Lookup Technique Using Distributed Memory Organization and ISCB-Tree Data Structure
The IP Lookup Process is a key bottleneck in routing due to the increase in routing table size, increasing traıc and migration to IPv6 addresses. The IP address lookup involves computation of the Longest Prefix Matching (LPM), which existing solutions such as BSD Radix Tries, scale poorly when traıc in the router increases or when employed for IPv6 address lookups. In this paper, we describe a ...
متن کاملA High Performance Parallel IP Lookup Technique Using Distributed Memory Organization and ISCB-Tree Data Structure
The IP Lookup Process is a key bottleneck in routing due to the increase in routing table size, increasing traıc and migration to IPv6 addresses. The IP address lookup involves computation of the Longest Prefix Matching (LPM), which existing solutions such as BSD Radix Tries, scale poorly when traıc in the router increases or when employed for IPv6 address lookups. In this paper, we describe a ...
متن کاملBandwidth Management in MPLS Networks
Two major limitations in legacy IP networking technologies are the impossibility of predicting the bandwidth available on a link at any given time and the difficulty of controlling the allocation of shared bandwidth. Addressing these problems requires a change in the management of networking resources. Multi Protocol Label Switching (MPLS) is the industry-standard approach to switching and forw...
متن کامل