OpenMP/Benchmarks - Revision history

Neteler: +compiler flags overview link

2013-09-15T16:43:19Z

+compiler flags overview link

← Older revision		Revision as of 16:43, 15 September 2013
Line 1:		Line 1:
	The idea of this wiki page is to explore which compiler and compiler flag combinations are most useful for speeding up computations without degrading data quality.		The idea of this wiki page is to explore which compiler and compiler flag combinations are most useful for speeding up computations without degrading data quality.

			For an compiler flags overview, see [https://doc.zih.tu-dresden.de/hpc-wiki/bin/view/Compendium/Compilers HPC Compendium: Compilers]

	== Neighborhood analysis ==		== Neighborhood analysis ==
Line 336:		Line 338:
	\| ?		\| ?
	\|}		\|}

			[[Category: Massive data analysis]]
			[[Category: Parallelization]]

2013-07-24T20:46:38Z

Neighborhood analysis

← Older revision		Revision as of 20:46, 24 July 2013
Line 6:		Line 6:

	Code and an automated run-script can be found at:		Code and an automated run-script can be found at:
	https://trac.osgeo.org/grass/browser/sandbox/soeren/benchmarks/neighborhood_openmp		: https://trac.osgeo.org/grass/browser/sandbox/soeren/benchmarks/neighborhood_openmp


	It's best to run it 4 times for each case, discard the first and average the next 3.		It's best to run it 4 times for each case, discard the first and average the next 3.

2013-07-24T20:46:12Z

Neighborhood analysis

← Older revision		Revision as of 20:46, 24 July 2013
Line 3:		Line 3:
	== Neighborhood analysis ==		== Neighborhood analysis ==

	Performance using OpenMP ~~and~~ different compilers. Code and an automated run-script can be found at:		* Performance test using OpenMP with different compilers and compiler options.

			Code and an automated run-script can be found at:
	https://trac.osgeo.org/grass/browser/sandbox/soeren/benchmarks/neighborhood_openmp		https://trac.osgeo.org/grass/browser/sandbox/soeren/benchmarks/neighborhood_openmp

2013-07-24T20:45:13Z

Neighborhood analysis

← Older revision		Revision as of 20:45, 24 July 2013
Line 3:		Line 3:
	== Neighborhood analysis ==		== Neighborhood analysis ==

	Performance using OpenMP and different compilers:		Performance using OpenMP and different compilers. Code and an automated run-script can be found at:
	~~source\|svn~~/sandbox/soeren/benchmarks/neighborhood_openmp/		https://trac.osgeo.org/grass/browser/sandbox/soeren/benchmarks/neighborhood_openmp

	~~Best~~ to run it 4 times for each case, discard the first and average the next 3.		It's best to run it 4 times for each case, discard the first and average the next 3.

	Example usage:		Example usage:

2013-07-24T08:25:38Z

← Older revision		Revision as of 08:25, 24 July 2013
Line 21:		Line 21:
	time ./neighbor 5000 5000 23		time ./neighbor 5000 5000 23
	</source>		</source>

			=== Results table ===

	Test setup: 5000x5000 array with a window size of 23x23 cells		Test setup: 5000x5000 array with a window size of 23x23 cells

	~~=== Results table ===~~

	{\| border="1" class="wikitable sortable" style="margin: 1em 1em 1em 0; background: #f9f9f9; border: 1px #aaaaaa solid; border-collapse: collapse;"		{\| border="1" class="wikitable sortable" style="margin: 1em 1em 1em 0; background: #f9f9f9; border: 1px #aaaaaa solid; border-collapse: collapse;"

2013-07-24T08:23:14Z

Results table: icc i7

@@ Line 266: / Line 266: @@
 |Ubuntu 12.04.2 LTS
 |15.5 gb
 |}

2013-07-24T08:18:15Z

Results table: i7 3770 gcc

@@ Line 146: / Line 146: @@
 |Debian GNU/Linux 6.0.7 (squeeze)
 |8.0 gb
 |}

2013-07-24T08:02:22Z

Results table: -O3

@@ Line 98: / Line 98: @@
+|
 |-
 |}

2013-07-24T08:00:41Z

why

← Older revision		Revision as of 08:00, 24 July 2013
Line 1:		Line 1:
			The idea of this wiki page is to explore which compiler and compiler flag combinations are most useful for speeding up computations without degrading data quality.

	== Neighborhood analysis ==		== Neighborhood analysis ==

2013-07-24T07:55:18Z

Results table