amd吧 关注:792,888贴子:18,226,487
  • 19回复贴,共1

各位来测一下内存在不同利用率下的延迟吧

只看楼主收藏回复

偶然间看到一篇nga上的文章
(内存延迟和内存带宽利用率之间的关系 http://nga.178.com/read.php?tid=21282980&forder_by=postdatedesc&rand=239
我就下载了Intel® Memory Latency Checker v3.9 自己测试一下

附上aida64的测试图

(数据不好看可能是因为我正开着游戏呢,内存使用5.2/11.9 GB)
有没有闲得**的测一测发上来看看


IP属地:广东1楼2020-08-22 21:33回复
    intel官方下载(https://software.intel.com/content/www/us/en/develop/articles/intelr-memory-latency-checker.html
    要用管理员权限运行cmd或powershell来执行


    IP属地:广东2楼2020-08-22 21:35
    回复
      延迟为什么这么贵,频率太低?


      IP属地:广东来自Android客户端5楼2020-08-22 22:15
      收起回复
        3800X 3733 C18
        Intel(R) Memory Latency Checker - v3.9
        Measuring idle latencies (in ns)...
        Numa node
        Numa node 0
        0 77.5
        Measuring Peak Injection Memory Bandwidths for the system
        Bandwidths are in MB/sec (1 MB/sec = 1,000,000 Bytes/sec)
        Using all the threads from each core if Hyper-threading is enabled
        Using traffic with the following read-write ratios
        ALL Reads : 37299.6
        3:1 Reads-Writes : 27687.4
        2:1 Reads-Writes : 26313.8
        1:1 Reads-Writes : 23161.3
        Stream-triad like: 32341.9
        Measuring Memory Bandwidths between nodes within system
        Bandwidths are in MB/sec (1 MB/sec = 1,000,000 Bytes/sec)
        Using all the threads from each core if Hyper-threading is enabled
        Using Read-only traffic type
        Numa node
        Numa node 0
        0 37349.8
        Measuring Loaded Latencies for the system
        Using all the threads from each core if Hyper-threading is enabled
        Using Read-only traffic type
        Inject Latency Bandwidth
        Delay (ns) MB/sec
        ==========================
        00000 474.26 39399.7
        00002 475.48 39277.8
        00008 498.45 39729.0
        00015 488.27 39872.3
        00050 478.48 39891.6
        00100 450.51 39880.4
        00200 106.32 32834.7
        00300 95.29 23553.2
        00400 93.23 18261.0
        00500 88.91 15052.6
        00700 86.75 11091.0
        01000 82.43 8133.4
        01300 82.12 6464.0
        01700 81.81 5150.8
        02500 81.53 3767.9
        03500 81.54 2917.6
        05000 81.32 2282.6
        09000 81.07 1621.9
        20000 80.77 1167.9
        Measuring cache-to-cache transfer latency (in ns)...
        Local Socket L2->L2 HIT latency 16.9
        Local Socket L2->L2 HITM latency 17.2


        IP属地:上海6楼2020-08-23 10:14
        收起回复
          数据太少了,有木有更多大佬测测数据发上来看看


          IP属地:广东8楼2020-08-23 22:06
          回复
            这个软件测试出来的带宽amd要比intel低一大截,不知道是bug还是本来就是这样


            IP属地:广东来自Android客户端10楼2020-08-23 22:10
            收起回复


              IP属地:广东11楼2020-08-23 22:15
              收起回复



                IP属地:广东12楼2020-08-30 11:42
                回复