Changes
Page history
Update results for 64C rome with larger working set.
authored
Jan 15, 2021
by
Jan Eitzinger
Show whitespace changes
Inline
Side-by-side
AMD-Rome-S2-M4-C64.md
View page @
74cde7fc
...
@@ -15,7 +15,7 @@
...
@@ -15,7 +15,7 @@
| Compiler | AMD clang |
| Compiler | AMD clang |
|----------|-------------------------------------------------------------------|
|----------|-------------------------------------------------------------------|
| Version | AMD clang version 10.0.0 (CLANG: AOCC_2.2.0-Build#93 2020_06_25) |
| Version | AMD clang version 10.0.0 (CLANG: AOCC_2.2.0-Build#93 2020_06_25) |
+
----------
+
-------------------------------------------------------------------
+
|
----------
|
-------------------------------------------------------------------
|
```
```
Optimizing flags:
```-Ofast -fnt-store=aggressive -std=c99 -fopenmp```
Optimizing flags:
```-Ofast -fnt-store=aggressive -std=c99 -fopenmp```
...
@@ -26,33 +26,33 @@ All results are in ```GB/s```.
...
@@ -26,33 +26,33 @@ All results are in ```GB/s```.
Summary results:
Summary results:
```
```
+-----------------------------------------------
-
+
+-----------------------------------------------+
| Single core | 3
5.83
(Triad)
|
| Single core | 3
4.28
(
S
Triad) |
| Memory domain | 43.
15
(Sum with 16 cores)
|
| Memory domain | 43.
30
(Sum with 16 cores) |
| Socket | 1
82.28 (Update
with 16 cores) |
| Socket | 1
71.57 (Sum
with 16 cores)
|
| Node |
414.1
8 (Update with 16 cores)
|
| Node |
338.8
8 (Update with 16 cores)|
+-----------------------------------------------
-
+
+-----------------------------------------------+
```
```
Results for scaling within a memory domain:
Results for scaling within a memory domain:
```
```
#nt Init Sum Copy Update Triad Daxpy STriad SDaxpy
#nt Init Sum Copy Update Triad Daxpy STriad SDaxpy
1 23.42 8.1
7
33.
1
4 32.
21 35.83 35.13
3
5
.2
1
3
5.34
1 23.42 8.1
6
33.4
7
32.
14 33.56 33.27
3
4
.2
8
3
3.88
2 23.4
3
16.
24
40.5
8
41.
31
40.
87
39.6
9
40.0
3
39.
2
5
2 23.4
2
16.
13
40.5
6
41.
25
40.
53
39.6
8
40.0
0
39.
3
5
3 23.44 24.04 40.6
4
39.
47
39.3
1
37.9
5
38.23 37.4
4
3 23.44 24.04 40.6
7
39.
39
39.3
7
37.9
2
38.23 37.4
0
4 23.42 31.5
2
40.0
2
38.
78
38.
53 37.03
37.
45
36.
51
4 23.42 31.5
1
40.0
0
38.
61
38.
42 36.90
37.
32
36.
36
5 23.4
1
36.4
6
39.6
5
38.
67
38.1
6
36.
84
37.1
8
36.
27
5 23.4
3
36.4
0
39.6
9
38.
42
38.1
4
36.
75
37.1
1
36.
15
6 23.4
2
35.
78
39.1
3
38.
34
37.4
9
36.1
2
36.5
7
35.6
2
6 23.4
4
35.
69
39.1
1
38.
05
37.4
4
36.1
0
36.5
1
35.
5
6
7 23.44 3
5.96
38.59 37.
69
36.8
5
35.4
8
3
6.00
35.0
2
7 23.44 3
6.00
38.59 37.
44
36.8
3
35.4
7
3
5.96
35.0
0
8 23.45 36.7
2
38.0
8
3
7.15 36.35 35.00
35.
53
34.
56
8 23.45 36.7
8
38.0
5
3
6.92 36.27 34.95
35.
44
34.
49
9 2
5.88
39.
38
39.1
3
38.
56
37.6
7
36.5
4
36.8
3
36.05
9 2
6.11
39.
46
39.1
1
38.
32
37.6
3
36.5
7
36.8
0
36.05
10 28.1
4
40.3
5
39.5
7
3
9.1
6 38.
53
37.6
3
37.
75
37.
12
10 28.
3
1 40.3
9
39.5
8
3
8.8
6 38.
42
37.
5
6 37.
66
37.
05
11 30.2
5
40.
67 40.00
39.
7
4 39.1
7
38.4
9
38.3
9
37.9
5
11 30.2
9
40.
70 39.99
39.4
7
39.1
2
38.4
6
38.3
8
37.9
2
12 32.
27
41.3
0
40.3
0 40.24
39.
72
39.
20
38.9
4
38.6
8
12 32.
30
41.3
6
40.3
1 39.91
39.
66
39.
17
38.9
3
38.6
5
13 34.
16 41.96
40.6
2
40.
96
40.2
1
39.
90
39.52 39.3
7
13 34.
22 42.05
40.6
4
40.
42
40.2
2
39.
85
39.52 39.3
5
14 3
5.96
42.
5
6 40.78 4
1.29
40.5
9
40.
42
39.9
3
39.8
9
14 3
6.03
42.
6
6 40.78 4
0.75
40.5
8
40.
36
39.9
2
39.8
5
15 37.
53
43.0
2
40.
78 41.49
40.8
1
40.7
7
40.
1
9 40.
20
15 37.
66
43.0
9
40.
80
40.8
8
40.7
9
40.
6
9 40.
14 40.19
16 39.
00
43.
15
40.
69 41.70
40.8
9
40.
90
40.2
2
40.3
8
16 39.
11
43.
30
40.
74 40.86
40.8
4
40.
85
40.2
1
40.3
4
```
```
...
@@ -61,110 +61,110 @@ Results for scaling across memory domains. Shown are the results for the number
...
@@ -61,110 +61,110 @@ Results for scaling across memory domains. Shown are the results for the number
Init:
Init:
```
```
#nm 1 2 3 4 5 6 7 8
#nm 1 2 3 4 5 6 7 8
1 23.42 46.80 70.
17
93.
39
116.
72 139.90
163.
25
186.
5
1
1 23.42 46.80 70.
20
93.
61
116.
81 140.21
163.
33
186.
7
1
2 23.4
3
46.8
1
70.
20
93.
40
116.9
1
140.
0
7 163.
24
186.
93
2 23.4
2
46.8
2
70.
17
93.
62
116.9
9
140.
3
7 163.
66
186.
86
3 23.44 46.8
3
70.2
4
93.
4
8 11
6.78
140.
02
163.6
3
186.
35
3 23.44 46.8
5
70.2
1
93.
6
8 11
7.03
140.
48
163.6
0
186.
99
4 23.42 46.83 70.1
4
93.6
4
11
6.82
140.
0
2 163.
03
186.
4
9
4 23.42 46.83 70.1
2
93.6
3
11
7.03
140.
2
2 163.
42
186.9
7
5 23.4
1
46.86 70.2
5
93.6
3
11
6.89
140.
0
5 163.
55
18
6.24
5 23.4
3
46.86 70.2
4
93.6
7
11
7.07
140.
4
5 163.
61
18
7.13
6 23.4
2
46.86 70.
19
93.
65
11
6.96
140.5
3
163.
64
18
6.85
6 23.4
4
46.86 70.
26
93.
72
11
7.12
140.
4
5 163.
88
18
7.24
7 23.44 46.8
6
70.30 93.
6
7 117.1
2
140.
58
163.9
8
18
6.95
7 23.44 46.8
8
70.30 93.7
4
117.1
6
140.
60
163.9
6
18
7.38
8 23.45 46.89 70.3
2
93.76 117.1
3
140.
42
163.94 187.2
7
8 23.45 46.89 70.3
3
93.76 117.1
9
140.
57
163.94 187.
4
2
9 2
5.88 51.73
77.6
2
103.4
6
129.3
4
155.
20
180.9
6
206.
69
9 2
6.11 52.20
77.6
1
103.4
7
129.3
2
155.
15
180.9
7
206.
74
10 28.1
4
56.28 84.
35
112.4
1
140.6
5
168.6
6
196.
57
224.
67
10 28.
3
1 56.28 84.
40
112.4
9
140.6
0
168.6
4
196.
69
224.
91
11 30.2
5
60.
49
90.7
4
12
0.87
151.3
0
181.4
5
211.
47
241.
65
11 30.2
9
60.
55
90.7
6
12
1.05
151.3
4
181.4
2
211.
60
241.
80
12 32.
27
64.4
9
96.
82
12
9.12
161.
07
193.
26
225.
11
257.
60
12 32.
30
64.4
8
96.
75
12
8.97
161.
31
193.
35
225.
73
257.
81
13 34.
16
68.3
9
102.5
8
136.
61
170.
89
204.
7
9 239.3
0
273.
49
13 34.
22
68.3
5
102.5
3
136.
58
170.
71
204.9
3
239.
1
3 273.
26
14 3
5.96
71.9
2
107.
87
14
3.59
179.8
1
21
5.6
1 25
0.92
287.
76
14 3
6.03
71.9
5
107.
94
14
4.08
179.8
5
21
6.0
1 25
1.73
287.
91
15 37.
53
75.1
0
112.
4
8 150.
01
18
7.76
225.
0
6 263.
0
2 300.
74
15 37.
66
75.
2
1 112.8
4
150.
47
18
8.07
225.6
2
263.2
3
300.
96
16 39.
00 78.07 117.07 155.87 194.85
234.
03
272.
51
31
2.25
16 39.
11 77.91 116.95 156.07 195.04
234.
14
272.
70
31
1.39
```
```
Sum:
Sum:
```
```
#nm 1 2 3 4 5 6 7 8
#nm 1 2 3 4 5 6 7 8
1 8.1
7
16.
2
8 24.
26
32.
33
40.
33
48.
4
6 56.48 64.
87
1 8.1
6
16.
1
8 24.
17
32.
29
40.
25
48.
3
6 56.48 64.
51
2 16.
24
32.
45
48.33 64.
36
80.4
5
96.
45
112.
2
4 128.
16
2 16.
13
32.
23
48.33 64.
57
80.4
7
96.
39
112.4
2
128.
20
3 24.04 48.
1
0 72.
43
96.
67
119.7
2
144.
96
167.8
1
19
2.6
6
3 24.04 48.0
8
72.
15
96.
03
119.
9
7 144.
10
167.8
5
19
1.7
6
4 31.5
2
62.5
6
94.
35
125.
6
8
86.72 186.48 218.07 252.1
9
4 31.5
1
62.5
1
94.
47
125.
9
8
157.17 188.51 219.76 250.7
9
5 36.4
6
7
3.01
109.2
0
14
6.27 182.4
3 218.
32
25
4.8
2 29
2.7
7
5 36.4
0
7
2.95
109.2
8
14
5.62 181.8
3 218.
18
25
3.9
2 29
0.2
7
6 35.
78
71.
5
5 107.
27
142.
58
1
2
8.
45
213.63 24
6.21 284.28
6 35.
69
71.
3
5 107.
10
142.
49
1
7
8.
26
213.63 24
8.47 283.31
7 3
5.96
71.9
6
107.
4
8 143.
37
179.
03
21
2.1
9 2
48
.6
1
28
7.04
7 3
6.00
71.9
3
107.8
2
143.
69
179.
91
21
4.7
9 2
50
.6
8
28
5.62
8 36.7
2
73.
41 109.78
146.
40
182.
22
21
8.01
255.
4
7 29
1.35
8 36.7
8
73.
53 110.07
146.
76
182.
97
21
9.33
255.
0
7 29
0.99
9 39.
38
78.
64
11
7.82
15
6
.3
2
19
5
.0
2
234.
33
27
3.41 311.02
9 39.
46
78.
91
11
8.11
15
7
.3
7
19
6
.0
8
234.
54
27
2.76 310.78
10 40.3
5
80.
60
120.
55
16
0.54 200.38 239.11 278.60 319.3
5
10 40.3
9
80.
79
120.
94
16
1.00 201.01 240.46 279.21 318.9
5
11 40.
6
7 81.
10
121.7
0
162.
13
20
1
.3
5
24
1.95 279.86 322.24
11 40.7
0
81.
38
121.
8
7 162.
07
20
2
.3
2
24
2.50 281.34 320.99
12 41.3
0
82.
5
2 123.
36
164.
1
7 20
4.71
245.
30
285.
33
32
6.99
12 41.3
6
82.
6
2 123.
74
164.7
1
20
5.44
245.
52
285.
50
32
5.30
13 4
1.96
83.
88
125.
24
16
6.52
208.
33
24
9.05
289.
74 330.07
13 4
2.05
83.
95
125.
76
16
7.21
208.
10
24
8.89
289.
02 329.41
14 42.
5
6 8
4.96
127.
11
16
8.45
211.13 252.
30
29
1.75 334.65
14 42.
6
6 8
5.17
127.
55
16
9.72
211.13 252.
78
29
3.64 333.37
15 43.0
2
85.
76
12
7.99 170.28 212.49 253.65 294.62
336.
5
8
15 43.0
9
85.
93
12
8.61 171.14 213.21 255.18 295.19
336.
3
8
16 43.
15
86.
15
12
8.82 170.71
21
2
.8
7
25
4.60
296.2
4
33
7
.8
8
16 43.
30
86.
33
12
9.25 171.57
21
3
.8
0
25
5.15
296.2
3
33
6
.8
6
```
```
Copy
Copy
```
```
#nm 1 2 3 4 5 6 7 8
#nm 1 2 3 4 5 6 7 8
1 33.
14 66.22 100.61 132.80 165.62 198.30 232.71 263.91
1 33.
47 65.72 98.72 134.01 166.29 197.55 228.82 259.89
2 40.5
8
8
0.93
121.
07
161.
23
20
1
.2
6
24
1.39 280.37
322.
1
2
2 40.5
6
8
1.02
121.
53
161.
79
20
2
.2
0
24
2.45 283.13
322.
6
2
3 40.6
4
81.2
9
12
1.72
162.
27
20
2.93
243.
4
6 28
3.27 323.31
3 40.6
7
81.2
4
12
2.03
162.
61
20
3.09
243.6
9
28
4.12 324.14
4 40.0
2
79.9
6
119.8
0
159.
40
19
7.78 238.99
278.
07
31
6.26
4 40.0
0
79.9
2
119.8
4
159.
79
19
9.55 239.05
278.
52
31
7.78
5 39.6
5
79.3
2
11
8.8
9 158.
29
19
8.21 238.02 278.26 317.75
5 39.6
9
79.3
0
11
9.0
9 158.
51
19
7.86 237.57 276.90 315.90
6 39.1
3
78.
24
117.
07
156.2
5
195.1
0
23
3.99 272.62 310.97
6 39.1
1
78.
11
117.
28
156.
3
2 195.1
7
23
4.26 273.25 311.62
7 38.59 77.1
8
115.7
0
15
3.83
192.
31
2
2
9.
83 268.74 305.98
7 38.59 77.1
4
115.7
1
15
4.25
192.
60 231.05
2
6
9.
28 307.35
8 38.0
8
76.
07
114.0
3
152.0
1
189.
34
227.
66
265.
25
30
2
.0
2
8 38.0
5
76.
15
114.0
9
152.0
9
189.
82
227.
73
265.
38
30
3
.0
9
9 39.1
3
78.2
3
117.
4
7 156.
45
195.
88
23
5.18 274.36 312.44
9 39.1
1
78.2
2
117.
3
7 156.
33
195.
20
23
3.89 272.85 311.33
10 39.5
7
79.1
3
118.
75
158.
06
197.
35
236.9
2
27
7.64 316.05
10 39.5
8
79.1
6
118.
62
158.
18
197.
61
236.
5
9 27
5.47 314.54
11
40.00
79.9
9
119.
8
5 159.
56
199.2
9
23
8.81 279.20 317.96
11
39.99
79.9
5
119.
9
5 159.
70
199.
5
2 23
9.03 278.54 318.17
12 40.3
0
80.5
6
120.
77
16
0.86
200.
73
240.1
3
280.
8
1 31
8.94
12 40.3
1
80.5
4
120.
89
16
1.01
200.
86
240.
7
1 280.
5
1 31
9.87
13 40.6
2
81.
31
12
2.19 163.75 205.94 247.65 289.42 331.40
13 40.6
4
81.
24
12
1.81 162.35 202.66 242.91 283.15 322.65
14 40.78 81.
48
122.
10
162.5
3
203.
86
24
5.19 288.91 335.03
14 40.78 81.
53
122.
24
162.
7
5 203.
25
24
3.50 283.65 323.49
15 40.
7
8 81.5
4
122.
24
162.
58
203.
2
4 24
4.12 286.56 331.7
0
15 40.8
0
81.5
6
122.
30
162.
90
203.4
1
24
3.66 284.17 323.9
0
16 40.
69
81.3
2
12
1.87
162.5
4
20
2.72
243.
18
28
5.64 329.20
16 40.
74
81.3
8
12
2.09
162.5
6
20
3.08
243.
23
28
3.37 323.38
```
```
Update
Update
```
```
#nm 1 2 3 4 5 6 7 8
#nm 1 2 3 4 5 6 7 8
1 32.
2
1 64.
50 96.82 129.53 162.49 193.57 224.44 259.21
1 32.1
4
64.
33 95.57 128.60 160.72 190.89 222.81 256.62
2 41.
31
82.
63
12
4.20 165.50 206.95 248.68 289.79 332
.6
4
2 41.
25
82.
58
12
3.52 164.81 205.70 247.46 289.00 329
.6
2
3 39.
47 79.20 119.26 159.33 200.45 241.61 280.88 323
.2
1
3 39.
39 78.76 118.38 158.02 197.89 237.18 276.75 316
.2
5
4 38.
78
77.
8
6 11
7.09 156.39 195.74 235.17 275.95 315
.9
9
4 38.
61
77.
2
6 11
6.14 154.66 193.32 232.05 270.46 308
.9
2
5 38.
67 78.07 118.46 158.73 200.38 242.20
2
8
1.
01 327.63
5 38.
42 76.98 115.74 154.28 193.22 232.44
2
7
1.
33 310.38
6 38.
34 77.22 116.79 157.32
19
8
.43 23
9.64 282.76 326.79
6 38.
05 76.15 114.57 152.86
19
1
.43 23
0.39 268.81 308.06
7 37.
69 76.09 115.36 154.97 195.94 237.15 279.56 322.33
7 37.
44 74.94 112.73 150.53 188.30 226.36 264.44 303.16
8 3
7.15
7
4
.9
5
11
3.25 152.32 192.08 232.98 273.96 315.53
8 3
6.92
7
3
.9
6
11
1.04 148.40 185.76 223.08 260.63 298.21
9 38.
56 77.88 117.97 158.68 200.08 242.66 284.03 328.92
9 38.
32 76.75 115.41 153.85 192.52 231.71 270.41 309.85
10 3
9.16 79.03 119.90 161.39 204.38 248
.1
9
2
92.73 338.33
10 3
8.86 77.88 117.03 156.31 195.53 235
.1
5
2
74.93 315.11
11 39.
74 80.39 121.76 164.08 207.84 251
.6
5
2
97.63 343.84
11 39.
47 79.05 118.84 158.64 198
.6
1
2
38.89 279.14 319.47
12
40.24 81.28
12
3
.1
9
16
6.43 210.07 254.30 300.00 346.2
0
12
39.91 79.93
12
0
.1
7
16
0.56 201.14 241.60 282.40 323.3
0
13 40.
96 83.53 127.5
9 1
7
3.
09 219.23 265.50 314.49 364.4
2
13 40.
42 81.06 121.9
9 1
6
3.
16 204.68 246.34 288.20 330.8
2
14 4
1.29 84.26 128.40
1
7
4.
20 222.00 272.58 325.20 374.6
3
14 4
0.75 81.69 122.97
1
6
4.
49 206.24 248.34 290.81 333.3
3
15 4
1.49 84.46 128.84
1
7
5.
51 225.82 277.54 335.09 395.36
15 4
0.88 82.01 123.41
1
6
5.
13 207.02 249.33 291.69 334.77
16 4
1.70
8
5
.9
7
1
32.67 182.28 235.50 290.26 348.92 414.1
8
16 4
0.86
8
1
.9
6
1
23.73 165.81 208.43 251.34 294.79 338.8
8
```
```
Triad
Triad
```
```
#nm 1 2 3 4 5 6 7 8
#nm 1 2 3 4 5 6 7 8
1 3
5.83 71.78 107.20 142.32 177.4
7 2
1
3.
48 247
.0
2
2
83.48
1 3
3.56 68.24 101.69 135.64 169.6
7 2
0
3.
60 238
.0
3
2
70.19
2 40.
87 81.62
12
2
.39 16
3.07 203.44 243.03 283.41
32
3
.4
0
2 40.
53 80.99
12
1
.39 16
1.72 202.04 242.32 282.53
32
2
.4
5
3 39.3
1
78.
59
118.
07
157.
10
196.
2
1 235.
74
27
3.83 313.68
3 39.3
7
78.
66
118.
11
157.
44
196.
6
1 235.
50
27
4.99 314.15
4 38.
53
76.
7
9 115.
3
0 153.2
6
191.3
5
229.
25
26
6.80 304.19
4 38.
42
76.9
1
115.
2
0 153.2
7
191.3
2
229.
47
26
7.65 305.44
5 38.1
6
76.
2
1 114.
58
152.4
5
190.
61
228.
50
266.
43
30
4.31
5 38.1
4
76.
3
1 114.
41
152.4
1
190.
44
228.
38
266.
20
30
3.83
6 37.4
9
74.8
4
112.
3
2 149.6
4
18
7.13
224.3
9
261.
21
298.
23
6 37.4
4
74.8
3
112.2
7
149.6
5
18
6.89
224.3
2
261.
48
298.
76
7 36.8
5
73.6
7
110.
37
147.0
5
183.
74
220.
05
256.5
1
292.
53
7 36.8
3
73.6
5
110.
44
147.
1
0 183.
69
220.
34
256.
9
5 292.
91
8 36.
35
72.6
7
108.
7
5 145.2
5
181.
45
217.
76
253.
14
288.
85
8 36.
27
72.6
5
108.
8
5 145.2
2
181.
10
217.
42
253.
03
288.
64
9 37.6
7
75.3
6
113.
14
150.4
5
18
8.15
225.
74
262.
60
299.
9
4
9 37.6
3
75.3
1
113.
03
150.4
7
18
7.96
225.
38
262.
41
299.
7
4
10 38.
53
76.
91
115.
50
15
4.15 192.40 230.79 269.53 307.39
10 38.
42
76.
86
115.
21
15
3.61 191.78 229.86 267.91 305.70
11 39.1
7
78.
37
117.
71
156.
74
19
6.09 235.00 273.80 312.87
11 39.1
2
78.
22
117.
32
156.
31
19
5.31 233.93 272.55 311.10
12 39.
72
79.
33
11
9.2
0 15
9.03 198.52 238.05 277.01 316.4
8
12 39.
66
79.
29
11
8.9
0 15
8.45 197.68 236.93 275.77 314.8
8
13 40.2
1
80.
50
120.
74
160.
92
200.
90
24
1.13 280.83 321.22
13 40.2
2
80.
43
120.
51
160.
61
200.
47
24
0.20 279.94 319.04
14 40.5
9
81.1
7
121.
82
16
2.15
202.
54
242.
53
282.
45
32
3.20
14 40.5
8
81.1
0
121.
57
16
1.97
202.
00
242.
27
282.
02
32
1.13
15 40.
81
81.
66
122.3
3
16
3.02 204.04 244.11 284.20 324.86
15 40.
79
81.
57
122.3
0
16
2.85 202.97 243.38 283.19 323.15
16 40.8
9
81.6
7
122.
67
16
3.15
203.
70
24
4.54 284.19 325.05
16 40.8
4
81.6
8
122.
42
16
2.89
203.
14
24
3.63 283.35 323.26
```
```
...
...
...
...