Intel Icelake S2 M1 C36
- Processor: Intel(R) Xeon(R) Platinum 8360Y CPU @ 2.40GHz
- Base frequency: 2.40GHz
- Number of sockets: 2
- Number of memory domains per socket: 1
- Number of cores per socket: 36
- Number of HWThreads per core: 2
- MachineState output: json
- Memory Info: 16x 16GB DDR4 3200MHz with ECC
+----------+-------------------------------+
| Compiler | icc (ICC) |
|----------|-------------------------------|
| Version | icc (ICC) 2021.2.0 20210228 |
+----------+-------------------------------+
Optimizing flags: -fast -xHost -qopt-streaming-stores=always -std=c99 -ffreestanding -qopenmp
All results are in GB/s.
Summary results:
+-----------------------------------------------+
| Single core | 26.75 (Init) |
| Memory domain | 173.34 (Sum with 35 cores) |
| Socket | 173.34 (Sum with 35 cores) |
| Node | 347.79 (Update with 35 cores)|
+-----------------------------------------------+
Results for scaling within a memory domain:
#nt Init Sum Copy Update Triad Daxpy STriad SDaxpy
1 26.75 15.20 24.87 14.23 20.18 20.93 19.49 19.69
2 53.43 29.07 47.84 27.10 39.01 40.50 37.82 37.86
3 76.53 42.20 61.98 38.53 54.94 57.47 53.29 52.95
4 101.71 56.19 80.58 49.23 72.78 74.03 70.70 68.28
5 126.19 68.18 97.19 59.88 88.59 87.92 86.61 81.84
6 146.44 78.42 109.36 69.17 101.48 100.47 99.57 93.55
7 152.58 89.87 119.96 78.63 113.55 112.50 111.14 105.14
8 154.12 97.77 125.41 86.37 121.12 119.62 120.34 113.45
9 155.14 107.09 130.20 93.98 127.62 127.15 128.17 121.62
10 155.81 114.33 134.19 102.48 132.98 133.00 133.28 128.38
11 153.89 121.99 139.71 107.94 138.78 137.73 139.56 135.46
12 154.71 129.35 141.17 116.63 141.63 140.63 143.43 139.25
13 154.53 135.84 144.36 121.05 144.53 144.24 146.62 143.13
14 152.50 140.15 148.03 128.28 147.54 147.30 149.27 145.82
15 150.08 141.60 148.93 134.33 149.42 147.38 150.67 147.27
16 150.12 149.28 151.99 141.54 151.89 152.47 152.03 151.43
17 148.89 151.01 152.30 145.70 154.58 152.92 155.27 154.32
18 144.85 154.63 154.40 147.88 155.58 156.79 155.59 155.24
19 145.05 154.47 155.62 150.91 156.94 157.09 156.41 155.68
20 143.51 157.83 157.41 152.13 154.98 154.96 155.37 154.70
21 142.51 160.68 158.94 156.54 158.35 157.50 158.42 156.55
22 142.56 159.28 158.69 154.97 158.19 158.34 158.58 157.62
23 139.98 163.21 159.75 158.34 158.06 159.16 158.25 158.03
24 139.49 165.35 161.94 160.03 160.02 160.53 160.44 159.44
25 137.31 164.06 160.61 158.20 159.96 160.41 159.87 158.75
26 136.74 167.28 163.18 163.01 160.70 162.14 160.82 160.49
27 133.57 168.35 161.64 163.36 160.00 161.57 160.56 160.25
28 133.93 169.60 162.55 163.12 159.95 161.15 160.88 160.13
29 135.81 169.84 161.66 166.30 160.87 162.94 160.66 161.19
30 134.02 170.38 161.53 166.07 159.83 161.99 160.72 160.11
31 134.49 171.64 162.37 165.07 159.70 161.45 160.71 160.05
32 131.46 172.86 161.65 166.84 159.98 162.60 160.64 160.79
33 131.09 171.90 160.39 165.57 158.68 161.51 159.65 159.33
34 129.17 173.22 161.33 167.59 158.96 161.15 159.94 159.70
35 130.03 173.34 161.18 165.50 158.03 160.03 159.68 158.54
36 130.58 172.92 160.48 166.15 159.12 160.72 159.95 159.22
Results for scaling across memory domains. Shown are the results for the number of memory domains used (nm) with columns number of cores used per memory domain.
Init:
#nm 1 2
1 26.75 52.97
2 53.43 105.49
3 76.53 151.17
4 101.71 200.21
5 126.19 247.48
6 146.44 274.82
7 152.58 278.28
8 154.12 285.23
9 155.14 285.60
10 155.81 280.11
11 153.89 278.66
12 154.71 276.52
13 154.53 278.45
14 152.50 278.37
15 150.08 274.81
16 150.12 272.96
17 148.89 270.42
18 144.85 272.24
19 145.05 266.50
20 143.51 268.18
21 142.51 260.83
22 142.56 262.93
23 139.98 265.38
24 139.49 261.44
25 137.31 258.86
26 136.74 261.53
27 133.57 256.21
28 133.93 251.81
29 135.81 258.73
30 134.02 250.55
31 134.49 253.48
32 131.46 254.84
33 131.09 251.98
34 129.17 254.49
35 130.03 249.36
36 130.58 247.51
Sum:
#nm 1 2
1 15.20 30.55
2 29.07 58.12
3 42.20 84.32
4 56.19 113.61
5 68.18 136.10
6 78.42 160.78
7 89.87 178.66
8 97.77 197.22
9 107.09 213.45
10 114.33 230.41
11 121.99 245.34
12 129.35 257.39
13 135.84 267.44
14 140.15 277.46
15 141.60 287.15
16 149.28 294.37
17 151.01 296.71
18 154.63 306.12
19 154.47 310.74
20 157.83 316.75
21 160.68 316.69
22 159.28 320.32
23 163.21 323.15
24 165.35 318.78
25 164.06 325.28
26 167.28 316.69
27 168.35 333.99
28 169.60 327.95
29 169.84 335.59
30 170.38 337.68
31 171.64 333.02
32 172.86 332.87
33 171.90 337.12
34 173.22 340.78
35 173.34 342.35
36 172.92 327.94
Copy
#nm 1 2
1 24.87 49.86
2 47.84 95.59
3 61.98 124.46
4 80.58 161.67
5 97.19 194.69
6 109.36 220.75
7 119.96 237.34
8 125.41 251.51
9 130.20 261.59
10 134.19 268.43
11 139.71 275.30
12 141.17 284.76
13 144.36 288.33
14 148.03 293.55
15 148.93 295.88
16 151.99 302.50
17 152.30 306.00
18 154.40 309.19
19 155.62 313.75
20 157.41 315.76
21 158.94 317.40
22 158.69 319.20
23 159.75 320.28
24 161.94 321.31
25 160.61 325.49
26 163.18 326.96
27 161.64 325.90
28 162.55 328.53
29 161.66 330.12
30 161.53 328.69
31 162.37 329.47
32 161.65 328.08
33 160.39 325.24
34 161.33 327.05
35 161.18 319.94
36 160.48 321.98
Update
#nm 1 2
1 14.23 28.51
2 27.10 54.33
3 38.53 77.31
4 49.23 99.54
5 59.88 120.47
6 69.17 141.09
7 78.63 159.26
8 86.37 173.95
9 93.98 192.40
10 102.48 204.09
11 107.94 217.88
12 116.63 236.11
13 121.05 248.22
14 128.28 258.11
15 134.33 273.55
16 141.54 284.83
17 145.70 290.22
18 147.88 297.88
19 150.91 308.80
20 152.13 314.96
21 156.54 318.96
22 154.97 323.65
23 158.34 321.88
24 160.03 326.76
25 158.20 331.79
26 163.01 336.16
27 163.36 337.13
28 163.12 337.61
29 166.30 338.44
30 166.07 341.45
31 165.07 336.71
32 166.84 342.55
33 165.57 343.25
34 167.59 339.70
35 165.50 347.79
36 166.15 325.54
Triad
#nm 1 2
1 20.18 40.32
2 39.01 78.08
3 54.94 109.47
4 72.78 145.31
5 88.59 176.18
6 101.48 203.73
7 113.55 225.24
8 121.12 239.75
9 127.62 254.59
10 132.98 263.17
11 138.78 273.30
12 141.63 282.42
13 144.53 289.16
14 147.54 292.37
15 149.42 297.28
16 151.89 301.95
17 154.58 305.79
18 155.58 307.34
19 156.94 309.99
20 154.98 311.17
21 158.35 316.45
22 158.19 315.71
23 158.06 316.41
24 160.02 316.33
25 159.96 316.76
26 160.70 318.57
27 160.00 319.24
28 159.95 315.35
29 160.87 319.29
30 159.83 317.69
31 159.70 318.24
32 159.98 317.61
33 158.68 317.72
34 158.96 315.32
35 158.03 315.14
36 159.12 313.57
Memory bandwidth scaling within one memory domain:
The following plots illustrate the the performance scaling over multiple memory domains using different number of cores per memory domain.
Memory bandwidth scaling across memory domains for init:
Memory bandwidth scaling across memory domains for sum
Memory bandwidth scaling across memory domains for copy
Memory bandwidth scaling across memory domains for Triad
