Home Go Up

NEC 5800

Machine Description

Model Number Express5800
Processor Type Itanium-2 @ 1.5GHz
Number of Processors 32 (31 used for FMS)
Memory Installed 128 Gbytes
Memory Used by FMS 64 Gbytes
Operating System Linux 2.4.18-nec3.2p1.033

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 100,000
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
3/09/04 20,000 Mem. 3.9 39:50 1:40 0 4,463 106,623 23.9
3/10/04 20,000 Disk 3.9 43:58 1:53 :06 4,044 94,070 23.3
3/10/04 40,000 Disk 14.9 5:01:36 11:15 :16 4,715 126,376 26.8
3/10/04 60,000 Disk 40.1 15:53:21 35:03 :27 5,035 136,932 27.2
3/10/04 80,000 Disk 64.3 1::13:25:42 1:18:59 1:17 5,066 144,047 28.4
3/09/04 100,000 Disk 114.4 2::23:53:49 2:33:17 03:51 5,151 144,978 28.1

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
3/10/04 20,000 Mem. 7.8 1:18:30 2:59 0 4,530 119,259 26.3
3/10/04 20,000 Disk 7.8 1:27:29 3:28 :13 4,065 102,626 25.2
3/10/04 40,000 Disk 29.8 9:56:49 56:19 :30 4,766 133,430 28.0
3/10/04 60,000 Disk 80.0 1::07:07:44 1:05:42 1:45 5,140 146,143 28.4
3/10/04 80,000 Disk 128.6 3::00:39:24 2:31:15 3:34 5,220 150,449 28.8
3/11/04 100,000 Disk 228.7 5::20:17:02 4:55:55 12:09 5,280 150,200 28.4

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.


Home Go Up
Copyright © Multipath Corporation