PC GAMESS v. 7.0.4 benchmarks and scalability on 21-node Pentium 4 Infiniband Linux cluster


Number of nodes used

Test 1, Wall clock time and relative speedup

Test 3, Wall clock time and relative speedup

Test 5, Wall clock time and relative speedup

Test 6, Wall clock time and relative speedup

1

4570.8

1.00

8892.9

1.00

9096.5

1.00

25640.7

1.00

2

2318.8

1.97

4400.9

2.02

4712.7

1.93

12954.5

1.98

3

1556.3

2.94

2925.1

3.04

3252.2

2.80

8711.3

2.94

4

1176.6

3.88

2200.9

4.04

2498.6

3.64

6591.9

3.89

5

949.0

4.82

1734.1

5.13

2068.0

4.40

5324.3

4.82

6

795.3

5.75

1452.4

6.13

1772.0

5.13

4476.8

5.73

7

687.5

6.65

1246.1

7.14

1564.3

5.82

3871.8

6.62

8

604.0

7.57

1089.1

8.17

1397.9

6.51

3414.5

7.51

9

541.3

8.44

969.4

9.17

1287.4

7.07

3068.6

8.36

10

489.9

9.33

878.7

10.12

1177.5

7.73

2786.3

9.20

11

448.7

10.19

793.3

11.21

1100.2

8.27

2554.1

10.04

12

413.9

11.04

735.9

12.08

1030.7

8.83

2357.2

10.88

13

384.2

11.90

679.4

13.09

974.1

9.34

2196.2

11.68

14

359.4

12.72

631.1

14.09

921.3

9.87

2055.8

12.47

15

338.3

13.51

592.7

15.00

882.8

10.30

1939.7

13.22

16

319.3

14.32

554.0

16.05

846.4

10.75

1837.0

13.96

17

302.2

15.13

522.6

17.02

829.6

10.97

1740.1

14.74

18

288.0

15.87

495.5

17.95

795.7

11.43

1661.4

15.43

19

274.2

16.67

467.7

19.01

774.6

11.74

1589.0

16.14

20

262.2

17.43

448.4

19.83

750.3

12.12

1525.3

16.81

 

Graphical representation of scalability



OS and hardware description


Intel Pentium 4 E Prescott (revision F41) 3.4 GHz processor, 800 MHz FSB, Intel SE7221BK1-E baseboard, RAM 4x512 MB DDR2 533 Kingston (dual channel), HDD 80 Gb SATA Seagate, HTT disabled in BIOS. 21 nodes, interconnect: Mellanox Infiniband HCA 8x PCI-e MT25208, Mellanox MTS2400 IB switch, OS Linux SUSE Profesional 9.3, IBGD-1.8.0, mvapich version 0.9.8 (vapi).  More information on this cluster (in Russian)



Tests description


Test 1, single-point direct DFT (B3LYP) energy plus gradient for medium-size system (623 basis functions). View image

Test 3, single-point direct MP2 energy for medium-size system (623 basis functions, the same system as one used for Test 1). View image

Test 5, single-point direct CASSCF(12,12) for medium-size system (retinal molecule, cc-pVDZ, 565 Cartesian basis functions) using ALDET code. View image

Test 6, single-point direct CIS energy plus gradient of first excited state of medium-size system (porphyrin molecule, cc-pVTZ (aug-cc on Nitrogens), 1130 Cartesian basis functions, D2h group). View image


Test comments


All tests were run in standard parallel mode using dynamic load balancing over p2p interface. Test 5 is the most communication intensive and would scale better for larger job. Wall clock times are given on master node in seconds.


We are grateful to Dr. Victor Datsyuk (Rostov State University, Russia) for providing access to cluster and helpful comments.

Copyright © 2007 by Alex A. Granovsky

Press to visit PC GAMESS v. 7.0.4 benchmarks and scalability on SKIF K-1000 (large Opteron Infiniband Linux cluster)

Press to visit PC GAMESS' eight core systems performance comparison page

Press to visit PC GAMESS' Woodcrest vs. Opteron performance comparison page

Press to visit PC GAMESS Pentium 4 family Xeon processor benchmarks page to compare the results of these benchmarks with those obtained on Xeon DP processors.

Press to visit PC GAMESS Pentium 4 family benchmarks page to compare the results of these benchmarks with those obtained on various Netburst (Pentium 4 and Pentium D) processors.

Press to visit the PC GAMESS vs. WinGamess performance comparison page to compare the results of these benchmarks with those obtained on older processors. Input files can be found there too.