Skip to Main Content
FREE TRIALS

Benchmarks Overview

Ansys benchmarks provide comprehensive and fair comparative information concerning the performance of Ansys solvers on available hardware platforms. The benchmarks can be used to compare performance of different hardware platforms when running Ansys solvers. To accomplish this goal, Ansys provides the suite of simulation cases that make up the Ansys benchmark suite to its hardware partners for their use in benchmarking hardware performance; then the data is reported to Ansys and compiled for this site. Cases are selected to represent typical usage and cover a range of mesh sizes and physical models.

Benchmarking Terminology

Ansys Fluent Benchmarks

This benchmark suite provides Ansys Fluent hardware performance data measured using sets of benchmark problems selected to represent typical usage. The Ansys Fluent benchmark cases range in size from a few hundred- thousand cells to more than 100 million cells. The suite contains both pressure-based (segregated and coupled) and density-based implicit solver cases using a variety of cell types and a range of physics. These cases are used by hardware vendors to benchmark their hardware systems and the resulting performance data is then included in these pages. For further descriptions of benchmark problems, please go to the benchmark page related to the case you are interested in.

Release 2021 R1 Test Cases

4-Stroke spray guided Gasoline Direct Injection model. The case has around 2 million Mixed cells and uses the Standard K-e Turbulence, DPM, Species Partially Premixed Combustion, dynamic mesh model and Pressure based segregated solver, Green Gauss node based, unsteady solver.


Number of cells 2,000,000
Cell type Mixed
Models Standard K-e Turbulence, DPM, Species Partially Premixed Combustion, dynamic mesh
Solver Pressure based segregated solver, Green Gauss node based, unsteady
 


Results

Results are listed alphabetically by platform. Download

Last updated:  2021-06-09
Platform Cores Machines Core Solver
Rating		 Core Solver
Speedup		 Core Solver
Efficiency

AMD white box, EPYC 75F3, 2.95GHz, Dual-socket, HDR InfiniBand, Using Intel MPI 2019 Update 8

Download
64
1
591.7808
64.000
100.00%
128
2
957.3407
103.535
80.89%
256
4
1416.3934
153.180
59.84%
512
8
1736.6834
187.819
36.68%

Intel white box, Platinum 9242, 48 cores, 2.3 GHz, Dual-socket, HDR InfiniBand, using Fluent AVX2 binary

Download
96
1
527.6336
96.000
100.00%
192
2
783.6735
142.585
74.26%
384
4
987.4286
179.657
46.79%
768
8
984.6154
179.145
23.33%
Boeing Landing gear analysis. The case has around 15 million Mixed cells and uses the realizable LES, Acoustics model and the Pressure based coupled solver, Least Squares cell based, Unsteady solver.

Number of cells 15,000,000
Cell type Mixed
Models LES, Acoustics
Solver Pressure based coupled solver, Least Squares cell based, Unsteady
 

Results

Results are listed alphabetically by platform. Download

Last updated:  2021-06-09
Platform Cores Machines Core Solver
Rating		 Core Solver
Speedup		 Core Solver
Efficiency
AMD white box, EPYC 75F3, 2.95GHz, Dual-socket, HDR InfiniBand, Using Intel MPI 2019 Update 8

Download
64
1
507.6381
64.000
100.00%
128
2
1035.9712
130.609
102.04%
256
4
2187.3418
275.767
107.72%
512
8
4523.5602
570.304
111.39%

Intel white box, Platinum 9242, 48 cores, 2.3 GHz, Dual-socket, HDR InfiniBand, using Fluent AVX2 binary

Download
96
1
564.7059
96.000
100.00%
192
2
1147.4104
195.060
101.59%
192
2
1147.4104
195.060
4101.59%
768
8
4385.7868
745.584
97.08%
1536
16
8074.7664
1372.710
89.37%
3072
32
12342.8571
2098.286
68.30%

Cavity flow in a centrifugal pump. The case has around 2 million hexahedral cells and uses the Realizable K-e Turbulence, Mixture Multiphase model and the Pressure based coupled solver, Least Squares cell based, steady solver.


Number of cells 2,000,000
Cell type hexahedral
Models Realizable K-e Turbulence, Mixture Multiphase
Solver Pressure based coupled solver, Least Squares cell based, steady
 

Results

Results are listed alphabetically by platform. Download

Last updated:  2021-06-09
 
Platform Cores Machines Core Solver
Rating		 Core Solver
Speedup		 Core Solver
Efficiency
AMD white box, EPYC 75F3, 2.95GHz, Dual-socket, HDR InfiniBand, Using Intel MPI 2019 Update 8

Download
64
1
7140.4959
64.000
100.00%
128
2
17024.6305
152.591
119.21%
256
4
37161.2903
333.075
130.11%
512
8
64000.0000
573.630
112.04%

Intel white box, Platinum 9242, 48 cores, 2.3 GHz, Dual-socket, HDR InfiniBand, using Fluent AVX2 binary

Download
96
1
8150.9434
96.000
100.00%
192
2
15292.0354
180.106
93.81%
384
4
28800.0000
339.200
88.33%
768
8
53169.2308
626.215
81.54%

Circulating fluidized bed. The case has around 2 million mixed cells and uses the Laminar, Eulerian Multiphase model and the Pressure based segregated solver, Least Squares cell based, unsteady solver.


Number of cells 2,000,000
Cell type Mixed
Models Laminar, Eulerian Multiphase
Solver Pressure based segregated solver, Least Squares cell based, unsteady
 

Results

Results are listed alphabetically by platform. Download

Last updated:  2021-06-09

 
Platform Cores Machines Core Solver
Rating		 Core Solver
Speedup		 Core Solver
Efficiency
AMD white box, EPYC 75F3, 2.95GHz, Dual-socket, HDR InfiniBand, Using Intel MPI 2019 Update 8

Download
64
1
6545.4545
64.000
100.00%
128
2
       12126.3158
118.568
92.63%
256
4
20945.4545
204.800
80.00%
512
8
33882.3529
331.294
64.71%

Intel white box, Platinum 9242, 48 cores, 2.3 GHz, Dual-socket, HDR InfiniBand, using Fluent AVX2 binary

Download
96
1
6400.0000
96.000
100.00%
192
2
12000.0000
180.000
93.75%
384
4
20945.4545
314.182
81.82%
768
8
32000.0000
480.000
62.50%

Release 19.0 Test Cases

Boeing Landing gear analysis. The case has around 15 million Mixed cells and uses the realizable LES, Acoustics model and the Pressure based coupled solver, Least Squares cell based, Unsteady solver.

Number of cells 15,000,000
Cell type Mixed
Models LES, Acoustics
Solver Pressure based coupled solver, Least Squares cell based, Unsteady
 

Results

Results are listed alphabetically by platform. Download

Last updated: 2019-02-26
Platform Cores Machines Core Solver
Rating		 Core Solver
Speedup		 Core Solver
Efficiency
Cray XC50  with Intel Skylake Gold 6148 processor, 2.4 GHz, Cray Linux Environment 6.0 update 07 based on SUSE 12 SP3 using Cray Aries network, using Fluent AVX2 binary

Download
 
40
1
256.0759
40.000
100.00%
80
2
506.1511
79.063
98.83%
160
4
1015.2761
158.590
99.12%
320
8
2042.5532
319.054
99.70%
480
12
3074.7331
480.285
100.06%
 640 16 4056.3380 633.615 99.00%
 960  24 5333.3333 833.086 86.78%
 1280  32 7646.0177 1194.336 93.31%
 1920  48 10936.7089 1708.354 88.98%
 2560  64  13292.3077 2076.307 81.11%
3840  96  15157.8947 2367.719 61.66%
5120   128 16301.8868 2546.415 49.73% 
5760  144 17632.6531 2754.285 47.82%
 HPE ProLiant DL360 Gen9 with Intel E5-2697A v4 processor, 2.60 GHz, RHEL 7.4, Mellanox ConnectX-5 EDR 100Gb/s InfiniBand Adapters using MLNX_OFED 4.3 driver with Mellanox HPC-X 2.2 MPI (NOT provided with Fluent)

Download
32
1
177.9609
32.000
100.00%
64
2
358.2090
64.411
100.64%
128
4
729.7297
131.216
102.51%
256
8
1479.4521
266.027
103.92%
 1024
 32
5958.6207 
1071.448 
104.63% 
Cavity flow in a centrifugal pump. The case has around 2 million hexahedral cells and uses the Realizable K-e Turbulence, Mixture Multiphase model and the Pressure based coupled solver, Least Squares cell based, steady solver.

Number of cells 2,000,000
Cell type hexahedral
Models Realizable K-e Turbulence, Mixture Multiphase
Solver Pressure based coupled solver, Least Squares cell based, steady
 

Results

Results are listed alphabetically by platform. Download

Last updated: 2019-02-26
Platform Cores Machines Core Solver
Rating		 Core Solver
Speedup		 Core Solver
Efficiency
Cray XC50  with Intel Skylake Gold 6148 processor, 2.4 GHz, Cray Linux Environment 6.0 update 07 based on SUSE 12 SP3 using Cray Aries network, using Fluent AVX2 binary

Download
40
1
3425.1734
40.000
100.00%
80
2
6953.7223
81.207
101.51%
160
4
13824.0000
161.440
100.90%
320
8
26584.6154
310.462
97.02%
480
12
39724.1379
463.908
96.65%
 640 16 53169.2308 620.923 97.02%
 960  24 65207.5472 761.509 79.32%
 1280  32 72000.0000 840.833 65.69%
 1920  48 82285.7143 960.952 50.05%
 2560  64  93405.4054 1090.811 42.61%
3840  96  96000.0000 1121.111 29.20%
5120   128 96000.0000 1121.111 21.90% 
HPE ProLiant DL360 Gen9 with Intel E5-2697A v4 processor, 2.60 GHz, RHEL 7.4, Mellanox ConnectX-5 EDR 100Gb/s InfiniBand Adapters using MLNX_OFED 4.3 driver with Mellanox HPC-X 2.2 MPI (NOT provided with Fluent)

Download
128
4
9931.0345
128.000
100.00%
512
16
42666.6667
549.926
107.41%
1024
32
75130.4348
968.348
94.57%

Circulating fluidized bed. The case has around 2 million mixed cells and uses the Laminar, Eulerian Multiphase model and the Pressure based segregated solver, Least Squares cell based, unsteady solver.


Number of cells 2,000,000
Cell type Mixed
Models Laminar, Eulerian Multiphase
Solver Pressure based segregated solver, Least Squares cell based, unsteady
 

Results

Results are listed alphabetically by platform. Download

Last updated: 2019-02-26
Platform Cores Machines Core Solver
Rating		 Core Solver
Speedup		 Core Solver
Efficiency
Cray XC50  with Intel Skylake Gold 6148 processor, 2.4 GHz, Cray Linux Environment 6.0 update 07 based on SUSE 12 SP3 using Cray Aries network, using Fluent AVX2 binary

Download
40
1
3023.6220
40.000
100.00%
80
2
5847.7157
77.360
96.70%
160
4
11076.9231
146.538
91.59%
320
8
20571.4286
272.143
85.04%
480
12
26790.6977
354.419
73.84%
 640 16 33882.3529 448.235 70.04%
 960  24 42666.6667 564.444 58.80%
 1280  32 48000.0000 635.000 49.61%
 1920  48 54857.1429 725.714 37.80%
HPE ProLiant DL360 Gen9 with Intel E5-2697A v4 processor, 2.60 GHz, RHEL 7.4, Mellanox ConnectX-5 EDR 100Gb/s InfiniBand Adapters using MLNX_OFED 4.3 driver with Mellanox HPC-X 2.2 MPI (NOT provided with Fluent)

Download
64
2
4363.6364
64.000
100.00%
128
4
8470.5882
124.235
97.06%
256
8
16225.3521
237.972
92.96%
512
16
28097.5610
412.098
80.49%
1024 
 32
46080.0000 
675.840 
66.00% 
External Flow over a Formula-1 Race car. The case has around 140 million Hex-core cells and uses the realizable k-e turbulence model and the Pressure based coupled solver, Least Squares cell based, pseudo transient solver.

Number of cells 140,000,000
Cell type Hex-core
Models Realizable K-e Turbulence
Solver Pressure based coupled solver, Least Squares cell based
 

Results

Results are listed alphabetically by platform.Download

Last updated: 2018-11-05
Platform Cores Machines Core Solver
Rating		 Core Solver
Speedup		 Core Solver
Efficiency
Cray XC50  with Intel Skylake Gold 6148 processor, 2.4 GHz, Cray Linux Environment 6.0 update 07 based on SUSE 12 SP3 using Cray Aries network, using Fluent AVX2 binary

Download
80
2
94.1074
80.000
100.00%
160
4
190.3922
161.851
101.16%
320
8
377.9528
321.295
100.40%
480
12
562.8664
478.489
99.69%
 640 16 745.9529 634.129 99.08%
 960  24 1129.7810 960.418 100.04%
 1280  32 1500.0000 1275.13 99.62%
 1920  48 2222.5080 1889.338 98.40%
 2560  64  2931.2977 2491.874 97.34%
3840  96  4194.1748 3565.437 92.85%
5120   128 5358.1395 4554.914 88.96%
5760  144 6063.1579 5154.245 89.48%
HPE ProLiant DL360 Gen9 with Intel E5-2697A v4 processor, 2.60 GHz, RHEL 7.4, Mellanox ConnectX-5 EDR 100Gb/s InfiniBand Adapters using MLNX_OFED 4.3 driver with Mellanox HPC-X 2.2 MPI (NOT provided with Fluent) 

Download
64
2
63.6066
64.000
100.00%
128
4
128.7966
129.593
101.24%
256
8
255.3003
256.879
100.34%
512
16
509.4340
512.585
100.11%
1024 
 32
1017.3683 
1023.661 
99.97 
External Flow over aircraft wing. The case has around 14 million hexahedral cells and uses the realizable K-e Turbulence model and the Pressure based coupled solver, Least Squares cell based, steady solver.

Number of cells 14,000,000
Cell type Hexahedral
Models Realizable K-e Turbulence
Solver Pressure based coupled solver, Least Squares cell based, steady
 

Results

Results are listed alphabetically by platform. Download

Last updated:  2019-02-26
Platform Cores Machines Core Solver
Rating		 Core Solver
Speedup		 Core Solver
Efficiency
Cray XC50  with Intel Skylake Gold 6148 processor, 2.4 GHz, Cray Linux Environment 6.0 update 07 based on SUSE 12 SP3 using Cray Aries network, using Fluent AVX2 binary

Download
40
1
699.4536
40.000
100.00%
80
2
1381.8473
79.024
98.78%
160
4
2802.9197
160.292
100.18%
320
8
5779.2642
330.502
103.28%
480
12
8640.0000
494.100
102.94%
 640 16 11520.0000 658.800 102.94%
 960  24 17366.8342 993.166 103.45%
 1280  32 22153.8462 1266.923 98.98%
 1920  48 31418.1818 1796.727 93.58%
 2560  64  38831.4607 2220.674 86.75% 
3840  96  48676.0563 2783.662 72.49% 
5120   128 54000.0000  3088.125  60.31% 
5760   144 57600.0000  3294.000  57.19% 
HPE ProLiant DL360 Gen9 with Intel E5-2697A v4 processor, 2.60 GHz, RHEL 7.4, Mellanox ConnectX-5 EDR 100Gb/s InfiniBand Adapters using MLNX_OFED 4.3 driver with Mellanox HPC-X 2.2 MPI (NOT provided with Fluent) 

Download
32
1
468.1658
32.000
100.00%
128
4
1969.2308
134.601
105.16%
512
16
8618.4539
589.087
115.06%
1024
32
17108.9109
1169.426
114.20%

Ansys Mechanical Benchmarks

This benchmark suite provides Ansys Mechanical hardware performance data measured using sets of benchmark problems selected to represent typical usage. The Ansys Mechanical benchmark cases range in size from 650,000 degrees of freedom to over 14 million degrees of freedom. The suite contains both PCG and Sparse equation solver benchmark cases as well as covering a wide variety of analysis types including static, transient, modal, cyclic symmetry modal and harmonic. Additionally, both thermal and structural physics are covered. These cases are used by hardware vendors to benchmark their hardware systems and the resulting performance data is then included in these pages. For further descriptions of benchmark problems, please go to the benchmark page related to the case you are interested in.

Ansys Mechanical Benchmarks 2019 R2

Transient nonlinear structural analysis of a electronic ball grid array
Analysis Type Static Nonlinear Structural
Number of Degrees of Freedom 6,000,000
Equation Solver Sparse
Matrix Symmetric

Plots

Results

Results are listed alphabetically by platform. Download

Last updated: 2020-01-30
Platform V19sp-5 RATINGS
Ansys Mechanical APDL
Version		 CPUs or Cores Machines or Blocks Core Solver Rating Core Solver Speedup Core Solver Efficiency
Intel Corporation Intel Xeon Gold 6254 Wolf Pass SP, Intel Xeon Gold 6254 (Cascade Lake) CPU, 192 GB RAM, CentOS Linux release 7.6.1810, none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 6.234 N/A N/A
8 1 11.75 N/A N/A
16 1 18.06 N/A N/A
32 1 26.80 N/A N/A
36 1 27.66 N/A N/A
Intel Corporation Intel Xeon Platinum 8260 Wolf Pass SP, Intel Xeon Platinum 8260L (Cascade Lake) CPU, 192 GB RAM, CentOS Linux release 7.6.1810, none, DMP config, 48 Total CPU Cores, 0 GPUs 19.4 4 1 6.283 N/A N/A
8 1 11.78 N/A N/A
16 1 17.66 N/A N/A
32 1 24.35 N/A N/A
36 1 23.63 N/A N/A
48 1 27.45 N/A N/A
Intel Corporation Intel Xeon E5-2697 v4 Grantley EP, Intel Xeon E5-2697 v4 (Broadwell) CPU, 128 GB RAM, CentOS Linux release 7.6.1810 , none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 4.175 N/A N/A
8 1 7.380 N/A N/A
16 1 11.11 N/A N/A
32 1 16.41 N/A N/A
36 1 15.31 N/A N/A
Intel Corporation Intel Xeon E5-2699 v3 Wildcat Pass EP, Intel Xeon E5-2699 v3 (Haswell) CPU, 128 GB RAM, CentOS Linux release 7.6.1810 , none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 3.934 N/A N/A
8 1 6.950 N/A N/A
16 1 10.64 N/A N/A
32 1 14.72 N/A N/A
36 1 14.77 N/A N/A
Static structural analysis of an engine block without the internal components

Analysis Type Static Linear Structural
Number of Degrees of Freedom 14,200,000
Equation Solver PCG
Matrix Symmetric
 

Plots

Results

Results are listed alphabetically by platform. Download

Last updated: 2020-01-30
Platform V19cg-3 RATINGS
Ansys Mechanical APDL
Version		 CPUs or Cores Machines or Blocks Core Solver Rating Core Solver Speedup Core Solver Efficiency
Intel Corporation Intel Xeon Gold 6254 Wolf Pass SP, Intel Xeon Gold 6254 (Cascade Lake) CPU, 192 GB RAM, CentOS Linux release 7.6.1810, none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 125.7 N/A N/A
8 1 209.3 N/A N/A
16 1 417.2 N/A N/A
32 1 689.5 N/A N/A
36 1 620.2 N/A N/A
Intel Corporation Intel Xeon Platinum 8260 Wolf Pass SP, Intel Xeon Platinum 8260L (Cascade Lake) CPU, 192 GB RAM, CentOS Linux release 7.6.1810, none, DMP config, 48 Total CPU Cores, 0 GPUs 19.4 4 1 126.2 N/A N/A
8 1 210.6 N/A N/A
16 1 409.3 N/A N/A
32 1 635.3 N/A N/A
36 1 613.2 N/A N/A
48 1 761.2 N/A N/A
Intel Corporation Intel Xeon E5-2697 v4 Grantley EP, Intel Xeon E5-2697 v4 (Broadwell) CPU, 128 GB RAM, CentOS Linux release 7.6.1810 , none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 107.6 N/A N/A
8 1 173.3 N/A N/A
16 1 300.4 N/A N/A
32 1 487.3 N/A N/A
36 1 510.3 N/A N/A
Intel Corporation Intel Xeon E5-2699 v3 Wildcat Pass EP, Intel Xeon E5-2699 v3 (Haswell) CPU, 128 GB RAM, CentOS Linux release 7.6.1810 , none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 102.3 N/A N/A
8 1 161.0 N/A N/A
16 1 284.1 N/A N/A
32 1 451.2 N/A N/A
36 1 460.1 N/A N/A
Modal analysis of a transmission housing without the internal components

Analysis Type Modal Structural
Number of Degrees of Freedom 7,700,000
Equation Solver PCG
Matrix Symmetric
 

Plots

Results

Results are listed alphabetically by platform. Download

Last updated: 2020-01-30
Platform V19ln-1 RATINGS
Ansys Mechanical APDL
Version		 CPUs or Cores Machines or Blocks Core Solver Rating Core Solver Speedup Core Solver Efficiency
Intel Corporation Intel Xeon Gold 6254 Wolf Pass SP, Intel Xeon Gold 6254 (Cascade Lake) CPU, 192 GB RAM, CentOS Linux release 7.6.1810, none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 48.23 N/A N/A
8 1 124.9 N/A N/A
16 1 289.3 N/A N/A
32 1 496.6 N/A N/A
36 1 521.7 N/A N/A
Intel Corporation Intel Xeon Platinum 8260 Wolf Pass SP, Intel Xeon Platinum 8260L (Cascade Lake) CPU, 192 GB RAM, CentOS Linux release 7.6.1810, none, DMP config, 48 Total CPU Cores, 0 GPUs 19.4 4 1 51.07 N/A N/A
8 1 132.8 N/A N/A
16 1 306.3 N/A N/A
32 1 523.0 N/A N/A
36 1 475.0 N/A N/A
48 1 572.6 N/A N/A
Intel Corporation Intel Xeon E5-2697 v4 Grantley EP, Intel Xeon E5-2697 v4 (Broadwell) CPU, 128 GB RAM, CentOS Linux release 7.6.1810 , none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 56.59 N/A N/A
8 1 133.3 N/A N/A
16 1 253.4 N/A N/A
32 1 362.3 N/A N/A
36 1 378.1 N/A N/A
Intel Corporation Intel Xeon E5-2699 v3 Wildcat Pass EP, Intel Xeon E5-2699 v3 (Haswell) CPU, 128 GB RAM, CentOS Linux release 7.6.1810 , none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 52.47 N/A N/A
8 1 125.0 N/A N/A
16 1 237.6 N/A N/A
32 1 324.1 N/A N/A
36 1 333.7 N/A N/A

Static nonlinear thermal-electric coupled field analysis of a Pelletier cooling block.

Analysis Type Static Nonlinear Thermal-Electric Coupled Field
Number of Degrees of Freedom 650,000
Equation Solver Sparse
Matrix Non-symmetric (Unsymmetric)
 

Plots

Results

Results are listed alphabetically by platform. Download

Last updated: 2020-01-30
 
Platform V19sp-1 RATINGS
ANSYS Mechanical APDL
Version		 CPUs or Cores Machines or Blocks Core Solver Rating Core Solver Speedup Core Solver Efficiency Solution Rating Solution Speedup Solution Efficiency
Intel Corporation Intel Xeon Gold 6254 Wolf Pass SP, Intel Xeon Gold 6254 (Cascade Lake) CPU, 192 GB RAM, CentOS Linux release 7.6.1810, none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 274.4 N/A N/A      
8 1 484.8 N/A N/A      
16 1 786.9 N/A N/A      
32 1 1096.4 N/A N/A      
36 1 1145.9 N/A N/A      
Intel Corporation Intel Xeon Platinum 8260 Wolf Pass SP, Intel Xeon Platinum 8260L (Cascade Lake) CPU, 192 GB RAM, CentOS Linux release 7.6.1810, none, DMP config, 48 Total CPU Cores, 0 GPUs 19.4 4 1 278.6 N/A N/A      
8 1 488.4 N/A N/A      
16 1 770.7 N/A N/A      
32 1 1002.3 N/A N/A      
36 1 984.1 N/A N/A      
48 1 1015.3 N/A N/A      
Intel Corporation Intel Xeon E5-2697 v4 Grantley EP, Intel Xeon E5-2697 v4 (Broadwell) CPU, 128 GB RAM, CentOS Linux release 7.6.1810 , none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 176.6 N/A N/A      
8 1 302.6 N/A N/A      
16 1 491.2 N/A N/A      
32 1 775.6 N/A N/A      
36 1 767.3 N/A N/A      
Intel Corporation Intel Xeon E5-2699 v3 Wildcat Pass EP, Intel Xeon E5-2699 v3 (Haswell) CPU, 128 GB RAM, CentOS Linux release 7.6.1810 , none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 167.3 N/A N/A      
8 1 284.0 N/A N/A      
16 1 457.9 N/A N/A      
32 1 702.4 N/A N/A      
36 1 690.6 N/A N/A      
Steady state thermal analysis of a power supply module

Analysis Type Steady State Thermal
Number of Degrees of Freedom 5,300,000
Equation Solver JCG
Matrix Symmetric
 

Plots

Results

Results are listed alphabetically by platform. Download

Last updated: 2020-01-30
Platform V19cg-1 RATINGS
Ansys Mechanical APDL
Version		 CPUs or Cores Machines or Blocks Core Solver Rating Core Solver Speedup Core Solver Efficiency
Intel Corporation Intel Xeon Gold 6254 Wolf Pass SP, Intel Xeon Gold 6254 (Cascade Lake) CPU, 192 GB RAM, CentOS Linux release 7.6.1810, none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 112.5 N/A N/A
8 1 207.7 N/A N/A
16 1 384.9 N/A N/A
32 1 609.7 N/A N/A
36 1 614.9 N/A N/A
Intel Corporation Intel Xeon Platinum 8260 Wolf Pass SP, Intel Xeon Platinum 8260L (Cascade Lake) CPU, 192 GB RAM, CentOS Linux release 7.6.1810, none, DMP config, 48 Total CPU Cores, 0 GPUs 19.4 4 1 111.9 N/A N/A
8 1 202.3 N/A N/A
16 1 378.4 N/A N/A
32 1 592.2 N/A N/A
36 1 541.4 N/A N/A
48 1 676.1 N/A N/A
Intel Corporation Intel Xeon E5-2697 v4 Grantley EP, Intel Xeon E5-2697 v4 (Broadwell) CPU, 128 GB RAM, CentOS Linux release 7.6.1810 , none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 106.5 N/A N/A
8 1 187.1 N/A N/A
16 1 311.1 N/A N/A
32 1 427.1 N/A N/A
36 1 443.3 N/A N/A
Intel Corporation Intel Xeon E5-2699 v3 Wildcat Pass EP, Intel Xeon E5-2699 v3 (Haswell) CPU, 128 GB RAM, CentOS Linux release 7.6.1810 , none, DMP config, 36 Total CPU Cores, 0 GPUs 19.4 4 1 98.46 N/A N/A
8 1 172.9 N/A N/A
16 1 286.3 N/A N/A
32 1 369.4 N/A N/A
36 1 382.8 N/A N/A
 

 

Ansys Mechanical Benchmarks 19.0

Static structural analysis of an engine block without the internal components

Analysis Type Static Linear Structural
Number of Degrees of Freedom 14,200,000
Equation Solver PCG
Matrix Symmetric
 

Plots


Results

Results are listed alphabetically by platform. Download

Last updated: 2018-08-23

 

Platform V19cg-3 RATINGS
Ansys Mechanical APDL
Version		 CPUs or Cores Machines or Blocks Core Solver Rating Core Solver Speedup Core Solver Efficiency
Lenovo ThinkStation P520, W-2195 CPU, 256 GB GB RAM, Windows 10, None, DMP config, 18 Total CPU Cores, 0 GPUs

 19.0 18 1 319.3 N/A N/A
Lenovo ThinkStation P520, W-2195 CPU, 256 GB GB RAM, Windows 10, NVIDIA GP100, DMP config, 18 Total CPU Cores, 1 GPUs

 19.0 18 1 319.6 N/A N/A
Lenovo ThinkStation P920, 2x Gold 6154 CPU, 384 GB RAM, Windows 10 16299, none, DMP config, 36 Total CPU Cores, 0 GPUs

 19.0 36 1 668.2 N/A N/A
Intel Corporation Intel Xeon Gold 6148 Whitebox, Intel Xeon Gold 6148 CPU, 192 GB RAM, Oracle Linux Server 7.5, none, DMP config, 40 Total CPU Cores, 0 GPUs

 19.1 1 1 26.78 1.0000 100.00%
2 1 62.43 2.331 116.6%
4 1 119.0 4.442 111.1%
8 1 194.0 7.245 90.56%
16 1 376.3 14.05 87.82%
24 1 461.3 17.22 71.77%
32 1 585.0 21.84 68.26%
40 1 628.4 23.46 58.66%
Intel Corporation Intel Xeon Gold 6148 Whitebox, Intel Xeon Gold 6148 CPU, 3072 GB RAM, Oracle Linux Server 7.5, none, DMP config, 512 Total CPU Cores, 0 GPUs

 19.1 32 1 580.3 N/A N/A
64 2 964.3 N/A N/A
128 4 1337.5 N/A N/A
256 8 1618.0 N/A N/A
512 16 1421.1 N/A N/A
Intel Corporation Intel Xeon E5-2697 v4 whitebox, Intel Xeon E5-2697 v4 CPU, 128 GB RAM, Oracle Linux Server 7.5, none, DMP config, 36 Total CPU Cores, 0 GPUs

 19.1 1 1 26.47 1.000 100.0%
2 1 60.11 2.271 113.5%
4 1 111.1 4.195 104.9%
8 1 174.6 6.594 82.43%
16 1 301.4 11.38 71.15%
24 1 391.1 14.77 61.56%
32 1 494.6 18.68 58.38%
36 1 495.1 18.70 51.95%
Steady state thermal analysis of a power supply module

Analysis Type Steady State Thermal
Number of Degrees of Freedom 5,300,000
Equation Solver JCG
Matrix Symmetric
 

Plots


Results

Results are listed alphabetically by platform. Download

Last updated: 2018-08-28

 

Platform V19cg-1 RATINGS
Ansys Mechanical APDL
Version		 CPUs or Cores Machines or Blocks Core Solver Rating Core Solver Speedup Core Solver Efficiency
Lenovo ThinkStation P520, W-2195 CPU, 256 GB GB RAM, Windows 10, None, DMP config, 18 Total CPU Cores, 0 GPUs

 19.0 18 1 275.5 N/A N/A
Lenovo ThinkStation P520, W-2195 CPU, 256 GB GB RAM, Windows 10, NVIDIA GP100, DMP config, 18 Total CPU Cores, 1 GPUs

 19.0 18 1 579.1 N/A N/A
Lenovo ThinkStation P920, 2x Gold 6154 CPU, 384 GB RAM, Windows 10 16299, none, DMP config, 36 Total CPU Cores, 0 GPUs

 19.0 36 1 398.9 N/A N/A
Intel Corporation Intel Xeon Gold 6148 Whitebox, Intel Xeon Gold 6148 CPU, 192 GB RAM, Oracle Linux Server 7.5, none, DMP config, 40 Total CPU Cores, 0 GPUs

 19.1 1 1 44.27 1.0000 100.00%
2 1 78.57 1.775 88.74%
4 1 110.0 2.485 62.13%
8 1 203.3 4.593 57.41%
16 1 357.8 8.081 50.50%
24 1 489.0 11.04 46.02%
32 1 575.6 13.00 40.63%
40 1 630.2 14.23 35.59%
Intel Corporation Intel Xeon Gold 6148 Whitebox, Intel Xeon Gold 6148 CPU, 3072 GB RAM, Oracle Linux Server 7.5, none, DMP config, 512 Total CPU Cores, 0 GPUs

 19.1 32 1 577.9 N/A N/A
64 2 1061.4 N/A N/A
128 4 1862.1 N/A N/A
256 8 3483.9 N/A N/A
512 16 5400.0 N/A N/A
Intel Corporation Intel Xeon E5-2697 v4 whitebox, Intel Xeon E5-2697 v4 CPU, 128 GB RAM, Oracle Linux Server 7.5, none, DMP config, 36 Total CPU Cores, 0 GPUs

 19.1 1 1 38.31 1.000 100.0%
2 1 70.44 1.839 91.94%
4 1 102.6 2.678 66.94%
8 1 182.6 4.767 59.59%
16 1 310.5 8.103 50.65%
24 1 384.9 10.05 41.86%
32 1 456.7 11.92 37.25%
36 1 470.3 12.28 34.10%
 

 
Static structural analysis of a farm tractor rear axle assemble

Analysis Type Static Linear Structural
Number of Degrees of Freedom 12,300,000
Equation Solver PCG
Matrix Symmetric
 
 

Plots


Results

Results are listed alphabetically by platform. Download

Last updated: 2017-08-28

 

Platform V19cg-2 RATINGS
Ansys Mechanical APDL
Version		 CPUs or Cores Machines or Blocks Core Solver Rating Core Solver Speedup Core Solver Efficiency
Lenovo ThinkStation P520, W-2195 CPU, 256 GB GB RAM, Windows 10, None, DMP config, 18 Total CPU Cores, 0 GPUs

 19.0 18 1 364.4 N/A N/A
Lenovo ThinkStation P520, W-2195 CPU, 256 GB GB RAM, Windows 10, NVIDIA GP100, DMP config, 18 Total CPU Cores, 1 GPUs

 19.0 18 1 388.5 N/A N/A
Lenovo ThinkStation P920, 2x Gold 6154 CPU, 384 GB RAM, Windows 10 16299, none, DMP config, 36 Total CPU Cores, 0 GPUs

 19.0 36 1 426.2 N/A N/A
Intel Corporation Intel Xeon Gold 6148 Whitebox, Intel Xeon Gold 6148 CPU, 192 GB RAM, Oracle Linux Server 7.5, none, DMP config, 40 Total CPU Cores, 0 GPUs

 19.1 1 1 48.20 1.0000 100.00%
2 1 73.16 1.518 75.88%
4 1 148.4 3.078 76.94%
8 1 285.4 5.921 74.02%
16 1 469.1 9.731 60.82%
24 1 462.0 9.585 39.94%
32 1 602.9 12.51 39.09%
40 1 605.5 12.56 31.40%
Intel Corporation Intel Xeon Gold 6148 Whitebox, Intel Xeon Gold 6148 CPU, 3072 GB RAM, Oracle Linux Server 7.5, none, DMP config, 512 Total CPU Cores, 0 GPUs

 19.1 32 1 590.6 N/A N/A
64 2 792.7 N/A N/A
128 4 841.3 N/A N/A
256 8 879.8 N/A N/A
512 16 856.3 N/A N/A
Intel Corporation Intel Xeon E5-2697 v4 whitebox, Intel Xeon E5-2697 v4 CPU, 128 GB RAM, Oracle Linux Server 7.5, none, DMP config, 36 Total CPU Cores, 0 GPUs

 19.1 1 1 43.39 1.000 100.0%
2 1 74.14 1.709 85.45%
4 1 140.5 3.238 80.95%
8 1 241.5 5.566 69.57%
16 1 375.0 8.643 54.02%
24 1 376.6 8.681 36.17%
32 1 460.1 10.60 33.14%
36 1 415.2 9.569 26.58%

 

Benchmarking Terminology

Rating

Rating is the primary metric used to report performance results of the benchmarks. It is defined as the number of benchmarks that can be run on a given machine (in sequence) in a 24 hour period. It is computed by dividing the number of seconds in a day (86400 seconds) by the number of seconds required to run the benchmark. A higher rating means faster performance.

Speedup

Speedup is the ratio of wall-clock time required to complete a given calculation using a single processor to that of the equivalent calculation performed on a concurrent machine. Its value ranges from 1 to the number of processors used for the parallel run. When the speedup is equal to the number of processors used, the speedup is called perfect or linear. Sometimes the speedup will exceed the number of processors. This is referred to as super-linear speedup, and is often caused by the availability and use of larger amounts of "fast" memory (e.g. cache or local memory) compared to the single processor run.

Efficiency

Efficiency is the speedup normalized by the number of processors used, presented as a percentage. It indicates the overall utilization of the CPUs during a parallel calculation. An efficiency of 100% indicates that each CPU is completely occupied with computation during the course of a run, and corresponds to a linear speedup. An efficiency of 60% indicates that each CPU is performing useful computation only 60% of the time. The remaining time is spent waiting for other functions, such as parallel communication or work on other processors, to complete.

See What Ansys Can Do For You

Contact us today

* = Required Field

Thank you for reaching out!

We’re here to answer your questions and look forward to speaking with you. A member of our Ansys sales team will contact you shortly.

Footer Image