我有这个C ++代码：
将10000个初始化的学生生成一个容器。 按字母顺序对学生进行排序。 对学生进行排序并通过失败。 将结果输出到控制台。
为了提高效率，使用不同的容器类型（静态和非静态）来执行定时来保持＆amp;迭代10000名学生。
没有从文件中读取任何内容，所有学生数据都包含在代码中。
这是使用的两个CPU之间的基准比较，从结果中可以明显看出一个是新的，一个是旧的：
http://cpuboss.com/cpus/Intel-Core-i7-3770K-vs-AMD-Opteron-170#performance
以下是比较每个CPU执行时间的结果....任何想法为什么新CPU被旧cpu留下了？ ：
-------------------------------------------------------------------------
AMD Opteron 170 - STATIC VECTOR ( 10,000 students = 27.499 secs )
-------------------------------------------------------------------------
gen_students = 1250ms 1.25s
sort_students = 9953ms 9.953s
alpha_pass = 7937ms 7.937s
pass_fail = 8359ms 8.359s
-------------------------------------------------------------------------
i7-3770K@3.5GHz - STATIC VECTOR ( 10,000 students = 46.675 secs )
-------------------------------------------------------------------------
gen_students = 2184ms 2.184s
sort_students = 32713ms 32.713s
alpha_pass = 5164ms 5.164s
pass_fail = 6614ms 6.614s
-------------------------------------------------------------------------
AMD Opteron 170 - STATIC LIST ( 10,000 students = 32.515 secs )
-------------------------------------------------------------------------
gen_students = 890ms 0.89s
sort_students = 15875ms 15.875s
alpha_pass = 7765ms 7.765s
pass_fail = 7985ms 7.985s
-------------------------------------------------------------------------
i7-3770K@3.5GHz - STATIC LIST ( 10,000 students = 27.221 secs )
-------------------------------------------------------------------------
gen_students = 374ms 0.374s
sort_students = 17160ms 17.16s
alpha_pass = 4633ms 4.633s
pass_fail = 5054ms 5.054s
-------------------------------------------------------------------------
AMD Opteron 170 - VECTOR ( 10,000 students = 552.094 secs )
-------------------------------------------------------------------------
gen_students = 1235ms 1.235s
sort_students = 534765ms 534.765s
alpha_pass = 7750ms 7.75s
pass_fail = 8344ms 8.344s
-------------------------------------------------------------------------
i7-3770K@3.5GHz - VECTOR ( 10,000 students = 896.07 secs )
-------------------------------------------------------------------------
gen_students = 2200ms 2.2s
sort_students = 882435ms 882.435s
alpha_pass = 4696ms 4.696s
pass_fail = 6739ms 6.739s
-----------------------------------------------------------
AMD Opteron 170 - LIST ( 10,000 students = 787.984 secs )
-----------------------------------------------------------
gen_students = 906ms 0.906s
sort_students = 771422ms 771.422s
alpha_pass = 7844ms 7.844s
pass_fail = 7812ms 7.812s
-------------------------------------------------------------------------
i7-3770K@3.5GHz - LIST ( 10,000 students = 398.645 secs )
-------------------------------------------------------------------------
gen_students = 358ms 0.358s
sort_students = 388412ms 388.412s
alpha_pass = 4758ms 4.758s
pass_fail = 5117ms 5.117s
答案 0 :(得分：2)
Opteron 170（旧版）计算机的二级缓存是新计算机的两倍，这对内存密集型操作有很大影响。当两个访问项目彼此靠近时，缓存效果最为明显 - 就像它们使用矢量一样 - 这正是我们在这里看到的。