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

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

延迟为什么这么贵，频率太低？

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

数据太少了，有木有更多大佬测测数据发上来看看

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