Intel Haswell S2 M1 C14
- Processor: Intel(R) Xeon(R) CPU E5-2695 v3
- Base frequency: 2.3 GHz
- Number of sockets: 2
- Number of memory domains per socket: 1
- Memory domain specs: 4-channel DDR4-2133
- Number of cores per socket: 14
- Number of HWThreads per core: 2
- MachineState output: json
+----------+---------------------------------+
| Compiler | icc (ICC) |
|----------|---------------------------------|
| Version | icc (ICC) 19.0.5.281 20190815 |
+----------+---------------------------------+
Optimizing flags: -fast -xHost -qopt-streaming-stores=always -std=c99 -ffreestanding -qopenmp
All results are in GB/s.
Summary results:
+----------------------------------------------+
| Single core | 19.25 (Triad) |
| Memory domain | 62.71 (Sum with 14 cores) |
| Socket | 62.71 (Sum with 14 cores) |
| Node | 125.40 (Sum with 13 cores) |
+----------------------------------------------+
Results for scaling within a memory domain:
#nt Init Sum Copy Update Triad Daxpy STriad SDaxpy
1 19.04 11.81 19.02 14.95 19.25 17.00 19.07 17.82
2 37.49 20.31 32.65 26.78 33.15 30.06 33.28 31.40
3 45.33 30.19 42.59 37.73 42.47 40.31 43.23 41.89
4 45.43 39.89 48.05 44.23 48.61 46.49 49.44 47.96
5 45.67 48.90 51.19 49.28 52.15 50.75 53.21 52.16
6 45.95 55.74 51.57 50.57 53.54 53.20 54.78 54.24
7 45.88 59.61 51.45 49.20 54.46 53.73 55.29 54.91
8 46.17 61.76 51.82 50.44 54.97 54.52 55.94 55.76
9 46.10 62.40 52.07 51.02 55.67 55.34 56.22 56.30
10 45.75 62.52 51.96 50.45 55.72 55.29 55.98 56.17
11 46.20 62.50 51.93 50.55 55.68 55.17 56.02 56.36
12 46.02 62.35 52.09 50.15 55.39 55.35 55.60 55.89
13 45.95 62.69 51.90 49.70 55.18 54.95 55.18 55.73
14 46.00 62.71 51.86 50.21 55.08 54.87 55.42 55.43
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 19.04 37.80
2 37.49 74.60
3 45.33 90.08
4 45.43 90.59
5 45.67 91.14
6 45.95 91.17
7 45.88 91.78
8 46.17 91.42
9 46.10 91.97
10 45.75 92.11
11 46.20 92.06
12 46.02 92.16
13 45.95 91.39
14 46.00 91.46
Sum:
#nm 1 2
1 11.81 23.56
2 20.31 45.49
3 30.19 64.68
4 39.89 82.17
5 48.90 98.36
6 55.74 111.90
7 59.61 118.42
8 61.76 122.92
9 62.40 124.26
10 62.52 124.36
11 62.50 124.73
12 62.35 124.18
13 62.69 125.40
14 62.71 125.37
Copy
#nm 1 2
1 19.02 37.81
2 32.65 69.20
3 42.59 84.29
4 48.05 95.79
5 51.19 101.04
6 51.57 103.74
7 51.45 103.00
8 51.82 103.72
9 52.07 103.74
10 51.96 103.53
11 51.93 103.48
12 52.09 102.92
13 51.90 103.42
14 51.86 102.16
Update
#nm 1 2
1 14.95 30.03
2 26.78 57.40
3 37.73 77.11
4 44.23 90.51
5 49.28 98.57
6 50.57 101.49
7 49.20 102.44
8 50.44 102.35
9 51.02 103.12
10 50.45 102.29
11 50.55 102.13
12 50.15 99.93
13 49.70 101.60
14 50.21 100.89
Triad
#nm 1 2
1 19.25 38.41
2 33.15 68.97
3 42.47 85.07
4 48.61 96.38
5 52.15 103.87
6 53.54 107.79
7 54.46 109.38
8 54.97 110.57
9 55.67 111.20
10 55.72 111.18
11 55.68 110.29
12 55.39 109.83
13 55.18 110.89
14 55.08 109.78
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
