NEMO3D - Revision history

Shirley at 20:22, 14 July 2009

2009-07-14T20:22:58Z

2008-09-03T20:45:04Z

‎Instrumentation overhead

2008-09-03T19:50:44Z

‎Legend

2008-09-03T19:50:18Z

‎Legend

2008-09-03T19:48:07Z

‎Runtime Breakdown Charts

2008-09-03T19:38:29Z

NEMO3DPerformance moved to NEMO3D: this page is the starting place for NEMO3D.

2008-09-03T19:34:22Z

‎Chart Runtime Breakdown

2008-09-03T18:41:04Z

2008-09-03T18:32:05Z

2008-09-03T18:12:57Z

‎Instrumentation overhead

← Older revision		Revision as of 20:22, 14 July 2009
Line 1:		Line 1:
−	This is a summary of the performance evaluation of NEMO3D. Our initial focus was on finding hot spots in the code where most of the computational work is being done. In all ~~case~~ the NEMO3D benchmark_lanc_thin_no_strain (262144) with recomputed matrices was used. Overhead was calculated on 16 processors.	+	This is a summary of the performance evaluation of NEMO3D. Our initial focus was on finding hot spots in the code where most of the computational work is being done. In all cases the NEMO3D benchmark_lanc_thin_no_strain (262144) with recomputed matrices was used. Overhead was calculated on 16 processors.

	==Instrumentation overhead==		==Instrumentation overhead==

@@ Line 13: / Line 13: @@
 |NA
 |-
-|Routine+loops instrumentation
+|Routine+loops instrumented
 |392
 |5.4

← Older revision		Revision as of 19:50, 3 September 2008
Line 25:		Line 25:
	<font color=black>Entire Experiment</font> <font color=red>Loop 1</font> <font color=green>Loop 2</font>		<font color=black>Entire Experiment</font> <font color=red>Loop 1</font> <font color=green>Loop 2</font>

−	[[Image:Nemo_plot_time.png\|~~500px~~\|left]] [[Image:Nemo_plot_fp.png\|~~500px~~\|left]] [[Image:Nemo_plot_dcm.png\|~~500px~~\|left]]	+	[[Image:Nemo_plot_time.png\|600px\|left]] [[Image:Nemo_plot_fp.png\|600px\|left]] [[Image:Nemo_plot_dcm.png\|600px\|left]]

← Older revision		Revision as of 19:50, 3 September 2008
Line 25:		Line 25:
	<font color=black>Entire Experiment</font> <font color=red>Loop 1</font> <font color=green>Loop 2</font>		<font color=black>Entire Experiment</font> <font color=red>Loop 1</font> <font color=green>Loop 2</font>

−	[[Image:Nemo_plot_time.png\|500px]] [[Image:Nemo_plot_fp.png\|500px]] [[Image:Nemo_plot_dcm.png\|500px]]	+	[[Image:Nemo_plot_time.png\|500px\|left]] [[Image:Nemo_plot_fp.png\|500px\|left]] [[Image:Nemo_plot_dcm.png\|500px\|left]]

← Older revision		Revision as of 19:48, 3 September 2008
Line 21:		Line 21:
	We have found two loops in the source file "h_cvectr_multi.c" that together account for about 90% (with 16 processors) of the runtime of the NEMO3D application. Loop 1 starts at 1235 and ends at 1841. Loop 2 starts at 1270 and ends at 1760.		We have found two loops in the source file "h_cvectr_multi.c" that together account for about 90% (with 16 processors) of the runtime of the NEMO3D application. Loop 1 starts at 1235 and ends at 1841. Loop 2 starts at 1270 and ends at 1760.
	===Runtime Breakdown Charts===		===Runtime Breakdown Charts===
		+	These charts show the runtime breakdown of NEMO3D on processor counts 16,32,64,128,256. Each experiment was run on PSC's SGI Altix system (pople).
		+	====Legend====
		+	<font color=black>Entire Experiment</font> <font color=red>Loop 1</font> <font color=green>Loop 2</font>
		+
		+	[[Image:Nemo_plot_time.png\|500px]] [[Image:Nemo_plot_fp.png\|500px]] [[Image:Nemo_plot_dcm.png\|500px]]

← Older revision	Revision as of 19:38, 3 September 2008
(No difference)

@@ Line 20: / Line 20: @@
 ==Significant Loops==
 We have found two loops in the source file "h_cvectr_multi.c" that together account for about 90% (with 16 processors) of the runtime of the NEMO3D application. Loop 1 starts at 1235 and ends at 1841. Loop 2 starts at 1270 and ends at 1760.
-===Chart Runtime Breakdown===
+===Runtime Breakdown Charts===

@@ Line 1: / Line 1: @@
-This is a summary of the performance evaluation of NEMO3D. Our initial focus was on finding hot spots in the code where most of the computational work is being done. In all case the NEMO3D benchmark
+This is a summary of the performance evaluation of NEMO3D. Our initial focus was on finding hot spots in the code where most of the computational work is being done. In all case the NEMO3D  benchmark_lanc_thin_no_strain (262144) with recomputed matrices was used. Overhead was calculated on 16 processors.
 ==Instrumentation overhead==
 {|
@@ Line 19: / Line 18: @@
 |}
+==Significant Loops==
 We have found two loops in the source file "h_cvectr_multi.c" that together account for about 90% (with 16 processors) of the runtime of the NEMO3D application. Loop 1 starts at 1235 and ends at 1841. Loop 2 starts at 1270 and ends at 1760.
-We have found two loops in the source file "h_cvectr_multi.c" that together account for about 90% (with 16 processors) of the runtime of the NEMO3D application. Loop 1 starts at 1235 and ends at 1841. Loop 2 starts at 1270 and ends at 1760.
+===Chart Runtime Breakdown===

← Older revision		Revision as of 18:12, 3 September 2008
Line 2:		Line 2:

	==Instrumentation overhead==		==Instrumentation overhead==
		+	Measured NEMO
		+
		+	{\|
		+	\|-
		+	! Run Type
		+	! Runtime (seconds)
		+	! Overhead %
		+	\|-
		+	\|Uninstrumented runtime
		+	\|372
		+	\|NA
		+	\|-
		+	\|Routine+loops instrumentation
		+	\|392
		+	\|5.4
		+	\|}
		+


	We have found two loops in the source file "h_cvectr_multi.c" that together account for about 90% (with 16 processors) of the runtime of the NEMO3D application. Loop 1 is		We have found two loops in the source file "h_cvectr_multi.c" that together account for about 90% (with 16 processors) of the runtime of the NEMO3D application. Loop 1 is