MatrixWarrior Powered by FMSlib Copyright © Multipath Corporation
Performance Analysis for Subroutine CNDF ; Time used=1:02:11
COMPLETED
20% I/O 54% MM 16% TR
Total Processor Performance
(64-bit floating point operations per second)
Tera   Giga   Mega   Kilo   Flops
  2 6   4 2 6   9 3 5   2 2 3   0 9 2
TUNING: See the Parameters page for tuning suggestions.

Component Performance (Gflops)
Routine All 40
CPUs		 GPU 1
40 °C
1245 MHz
39 Watts 		GPU 2
33 °C
1245 MHz
34 Watts 		GPU 3
34 °C
1245 MHz
35 Watts 		GPU 4
34 °C
1245 MHz
35 Watts
Matrix Multiply 27503 1628 6534 6639 6650 6643
CPU(4%)
Triangle Solve 25065 1221 6042 6072 6079 6070
CPU(3%)
Diagonal Factor 15869 7 4120 4141 4077 4069
GPU model = Tesla V100-PCIE-16GB
Time (Sec.)
Routine All CPUs GPU 1 GPU 2 GPU 3 GPU 4
Matrix Multiply 1998 1685 1997 1966 1962 1964
Triangle Solve 601 450 600 597 596 597
Diagonal Factor 126 7 120 120 123 124
Total Compute 2725 2142 2717 2683 2682 2686
Assembly 13
I/O wait 748
HTML 6
Other 253
Overall 3731
Times and Problem
Job Started 0 15:42:53 Wed Apr 11 2018 Equations 300000 Compute Rate (Gflops) 26427
Routine Started 111 15:44:44 Wed Apr 11 2018 Vectors 0 Read Rate (MBytes/Sec.) 2257
Current Time 3843 16:46:56 Wed Apr 11 2018 Block Size 50176 X 50176 Disk->Memory Reuse 100352
Estimated
Completion		 3843 16:46:56 Wed Apr 11 2018 Data Type Complex Disk Speed /
Compute Speed		 1.07