首页> 外文OA文献 >A Flexible High-Bandwidth Low-Latency Multi-Port Memory Controller
【2h】

A Flexible High-Bandwidth Low-Latency Multi-Port Memory Controller

机译:灵活的高带宽低延迟多端口存储器控制器

代理获取
本网站仅为用户提供外文OA文献查询和代理获取服务,本网站没有原文。下单后我们将采用程序或人工为您竭诚获取高质量的原文,但由于OA文献来源多样且变更频繁,仍可能出现获取不到、文献不完整或与标题不符等情况,如果获取不到我们将提供退款服务。请知悉。

摘要

Multi-port memory controllers (MPMCs) have become increasingly important inmany modern applications due to the tremendous growth in bandwidth requirement.Many approaches so far have focused on improving either the memory accesslatency or the bandwidth utilization for specific applications. Moreover, theapplication systems are likely to require certain adjustments to connect withan MPMC, since the MPMC interface is limited to a single-clock and singledata-width domain. In this paper, we propose efficient techniques to improvethe flexibility, latency, and bandwidth of an MPMC. Firstly, MPMC interfacesemploy a pair of dual-clock dual-port FIFOs at each port, so any multi-clockmulti-data-width application system can connect to an MPMC without requiringextra resources. Secondly, memory access latency is significantly reducedbecause parallel FIFOs temporarily keep the data transfer between theapplication system and memory. Lastly, a proposed arbitration scheme, namelywindow-based first-come-first-serve, considerably enhances the bandwidthutilization. Depending on the applications, MPMC can be properly configured byupdating several internal configuration registers. The experimental results inan Altera Cyclone FPGA prove that MPMC is fully operational at 150 MHz andsupports up to 32 concurrent connections at various clocks and data widths.More significantly, achieved bandwidth utilization is approximately 93.2% ofthe theoretical bandwidth, and the access latency is minimized as compared toprevious designs.
机译:由于带宽需求的巨大增长,在许多现代应用中,多端口内存控制器(MPMC)变得越来越重要。到目前为止,许多方法都专注于提高特定应用程序的内存访问延迟或带宽利用率。此外,由于MPMC接口仅限于单时钟和单数据宽度域,因此应用系统可能需要进行某些调整才能与MPMC连接。在本文中,我们提出了有效的技术来提高MPMC的灵活性,延迟和带宽。首先,MPMC接口在每个端口上使用一对双时钟双端口FIFO,因此任何多时钟多数据宽度的应用系统都可以连接到MPMC,而无需额外的资源。其次,由于并行FI​​FO暂时保持了应用程序系统与内存之间的数据传输,因此大大降低了内存访问延迟。最后,提出的仲裁方案,即基于窗口的先到先服务,大大提高了带宽利用率。根据应用的不同,可以通过更新几个内部配置寄存器来正确配置MPMC。 Altera Cyclone FPGA的实验结果证明MPMC可以在150 MHz下完全工作,并且在各种时钟和数据宽度下最多支持32个并发连接。更重要的是,实现的带宽利用率约为理论带宽的93.2%,并且将访问延迟最小化为与以前的设计相比

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
代理获取

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号