PCM 内存系统研究综述

夏飞, 蒋德钧, 熊劲, 孙凝晖 - 计算机科学技术学报, 2015 - jcst.ict.ac.cn
… It is difficult for DRAM (dynamic random access memory) based … , write command latency is
reduced as RESET takes less timesimultaneously. For example, D-XOR accelerates single …

[PDF][PDF] 妆䉬㢢喽大ु

李炍埕 - 2016 - ee.cityu.edu.hk
… loss is caused by the random access nature or sub-optimality of … decoded simultaneously if
the SINR threshold is sufficiently low… Therefore, we drop the superscript j, and write the steady-…

基于数据中心互连的新型内存架构

江涛, 侯锐, 董建波, 柴琳, 田斌, 张立新… - 计算机科学技术学报, 2015 - jcst.ict.ac.cn
… and bus simultaneously; changing the cache policy from write-… or issuing sustained writes
generates pipelined parallel TLP re… 4: one with a random read/write working set larger than …

内存制约加速比模型及其对计算的影响

孙贤和, 鲁潇阳 - 计算机科学技术学报, 2023 - jcst.ict.ac.cn
… , no matter how many processors are used, the upper bound of speedup is . The speedup of
is under the assumption that there is no other parallel … (nonvolatile random access memory) …

[PDF][PDF] ISTANBUL TECHNICAL UNIVERSITY 击INSTITUTE OF SCIENCE AND TECHNOLOGY

AB Senyüz - 2005 - Citeseer
… problems faster by using multiple processors simultaneously. … run in parallel and communicate
by reading from and writing to a … Parallel Random Access Machine (PRAM) is well known …

[PDF][PDF] 巴团g 喹

S Owicki, A Agarwal - 1989 - bitsavers.org
… We also used an 8-processor trace of PERO, which was obtained from a parallel tracer that
… data are simultaneously in the caches of two or more processors, and then only on write

基于赛道存储的寄存器堆在GPU 上的性能优化

梁云, 王硕 - 计算机科学技术学报, 2016 - jcst.ict.ac.cn
… density compared with SRAM (static random access memory)[9]. … read or write requests
could be served at the same time, which means up to four banks can be accessed simultaneously

[PDF][PDF] E 团g 围

LJ Guibas, J Stolfi - 1989 - researchgate.net
… In order to get the desired bounds on the space and time … allowed to move only by parallel
translation, without rotation. In this … We do this on both sides of l simultaneously, with pointers …

大规模数据管理系统综述

吴冷冬, 袁立言, 犹嘉槐 - 计算机科学技术学报, 2015 - jcst.ict.ac.cn
write MapReduce functions over data stored in their parallel … synchronized across multiple
machines without the need of … to run simultaneously on multiple computers for parallelism. The …

一种面向多核处理器系统的I/O 敏感的动态调度策略

吕方, 崔慧敏, 王蕾, 刘磊, 武成岗, 冯晓兵… - 计算机科学技术 …, 2014 - jcst.ict.ac.cn
concurrent kernels (inserting nops), we can get different ranks of simultaneous I/O requests.
… The most important premise of FIOS is the discrepancy between read time and write time on …