Home Go Up

Bull

Machine Description

Model Number Bull NovaScale 5160
Processor Type Itanium-2 @ 1.5GHz
Number of Processors 16
Memory Installed 64 Gbytes
Memory Used by FMS 32 Gbytes
Operating System Linux 2.6.7-B64K.1.5
Disk Configuration XFS File System
(8) Drives
(8) FMS Stripes (NUMIOQ=8)
Direct I/O (FMS IOTYPE=2)

Problem Description

Code Used Example 11
Matrix Sparsity Full
Matrix Symmetry Symmetric and Nonsymmetric
Data Type 16-byte Complex
Number of Equations 20,000 to 150,000
Number of Vectors 1
Reduced Operation Algorithms None (IALGOR=0)

Results for Complex Symmetric Matrices

Date
Started		 Number of
Equations		 Storage Times (Day::Hr:Min:Sec) Megaflops Parallel
Speedup
Loc. Gbytes CPU Wall I/O Wait CPU Overall
2/09/05 20,000 Mem. 3.7 37:37 2:41 0 4,727 66,166 14.0
2/09/05 20,000 Disk 3.7 38:06 2:44 :01 4,666 64,895 13.9
2/09/05 40,000 Mem. 17.2 5:00:22 20:45 0 4,735 68,567 14.5
2/09/05 40,000 Disk 17.2 5:00:28 20:45 :05 4,738 68,564 14.5
2/09/05 60,000 Disk 47.6 16:45:44 1:07:09 :28 4,772 71,487 15.0
2/09/05 80,000 Disk 68.6 1::15:38:28 2:37:16 1:32 4,783 72,348 15.1
2/09/05 100,000 Disk 96.0 3::04:51:12 5:02:50 2:18 4,819 73,385 15.2
2/10/05 150,000 Disk 205.6 10::15:20:18 16:35:56 5:06 4,895 75,307 15.4

Results for Complex Nonsymmetric Matrices

Date
Started		 Number of
Equations		 Storage Times (Day::Hr:Min:Sec) Megaflops Parallel
Speedup
Loc. Gbytes CPU Wall I/O Wait CPU Overall
2/08/05 20,000 Mem. 7.3 1:13:07 4:56 0 4,863 72,122 14.8
2/08/05 20,000 Disk 7.3 1:13:52 5:03 :03 4,814 70,374 14.6
2/08/05 40,000 Disk 34.3 9:48:15 39:17 27 4,835 72,423 15.0
2/08/05 60,000 Disk 95.2 1::09:04:26 2:12:41 3:06 4,837 72,359 15.0
2/09/05 80,000 Disk 137.1 3::06:31:59 5:10:06 4:43 4,829 73,382 15.2
2/09/05 100,000 Disk 191.9 6::08:45:00 9:57:53 6:37 4,849 74,337 15.3

NOTES:

  1. CPU Time is the total amount of User and System time used by all processors.
  2. Wall Time is the elapsed time measured on a dedicated machine. This includes the time spent processing and any time spent waiting for I/O to complete.
  3. I/O Wait Time is the total time spent waiting for I/O to complete that is not overlapped by asynchronous I/O.
  4. Megaflops per Processor is the theoretical number of floating point operations performed, divided by the CPU time.
  5. Overall Megaflops is the total number of floating point operations performed, divided by the Wall Time.
  6. Parallel Speedup is overall increase in performance due to parallel processing.


HomeGo Up
Copyright © Multipath Corporation