论文部分内容阅读
计算机运行速度的提高不仅与处理器性能有关,还需解决内存方面的“存储墙”问题。美国伊利诺伊理工大学的研究人员提出了一种名为C-AMAT的数学模型,通过平行内存读取方式使计算机的缓存和存储系统并行处理,可极大地减少计算机在数据读取方面的延迟,最多可使数据读取速度提高100倍。C-AMAT模型包括5个参数,分别表征命中存储请求的并发度、缺失存储请求的并发度、存储请求
The increase of the speed of the computer not only relates to the processor performance, but also needs to solve the memory problem of “storage wall”. Researchers at the University of Illinois at Chicago put forward a mathematical model called C-AMAT, which parallelizes the parallel processing of computer cache and storage systems with parallel memory reads, greatly reducing the computer latency in data reads, up to Can make data read speed 100 times. The C-AMAT model includes five parameters, which respectively represent the concurrency of the hit storage request, the concurrency of the missing storage request, the storage request