PDR.TetGen - Revision history

Ctsolakis: link to reproducibility

2018-03-29T09:50:38Z

link to reproducibility

← Older revision		Revision as of 05:50, 29 March 2018
Line 6:		Line 6:
	:# '''Code re-use''': a modular design of the parallel software that builds upon a previously designed sequential meshing code, such that it can be replaced and/or updated with a minimal effort. Due to the complexity of meshing codes, this is the only practical approach for keeping up with the ever-evolving sequential algorithms.		:# '''Code re-use''': a modular design of the parallel software that builds upon a previously designed sequential meshing code, such that it can be replaced and/or updated with a minimal effort. Due to the complexity of meshing codes, this is the only practical approach for keeping up with the ever-evolving sequential algorithms.
	:# '''Scalability''': the ratio of the time taken by the best sequential implementation to the time taken by the parallel implementation. The speedup is always limited by the inverse of the sequential fraction of the software, and therefore all non-trivial stages of the computation must be parallelized to leverage the current architectures with millions of cores.		:# '''Scalability''': the ratio of the time taken by the best sequential implementation to the time taken by the parallel implementation. The speedup is always limited by the inverse of the sequential fraction of the software, and therefore all non-trivial stages of the computation must be parallelized to leverage the current architectures with millions of cores.
	:# ''' Reproducibility ''': (weak & strong) TetGen meets none of these~~. For details see [LINK]~~. When it comes to shared memory no problem , in distributed memory ~~(explain)~~ This is why results are on limited geometries.		:# ''' Reproducibility ''': (weak & strong) TetGen meets none of these. When it comes to shared memory there is no problem since data are directly accessible. However, in distributed memory the procedure of packing and populated TetGen's data structures with an existing mesh creates invalid meshes see [[Reproducibility TetGen \| here]]. This is why results are on limited geometries.
	The design and implementation of a sequential industrial strength code is labor intensive, it takes about 100 man-years. Parallel mesh generation code is even more labor intensive (by an order of magnitude for traditional parallel machines and expected to be higher for current and emerging architectures due to multiple memory and network hierarchies, fault-tolerance and power aware issues); however, because of the underlying theory that allows code re-use, this initial parallel implementation was achieved in less than six months with impressive functionality, i.e., the same as the sequential TetGen code.		The design and implementation of a sequential industrial strength code is labor intensive, it takes about 100 man-years. Parallel mesh generation code is even more labor intensive (by an order of magnitude for traditional parallel machines and expected to be higher for current and emerging architectures due to multiple memory and network hierarchies, fault-tolerance and power aware issues); however, because of the underlying theory that allows code re-use, this initial parallel implementation was achieved in less than six months with impressive functionality, i.e., the same as the sequential TetGen code.

Ctsolakis: Ctsolakis moved page Cores/PDR to PDR.TetGen: conform with PDR.AFLR

2018-03-29T05:25:05Z

Ctsolakis moved page Cores/PDR to PDR.TetGen: conform with PDR.AFLR

← Older revision	Revision as of 01:25, 29 March 2018
(No difference)

Ctsolakis: /* Overview */

2018-03-28T09:18:29Z

Overview

← Older revision		Revision as of 05:18, 28 March 2018
Line 6:		Line 6:
	:# '''Code re-use''': a modular design of the parallel software that builds upon a previously designed sequential meshing code, such that it can be replaced and/or updated with a minimal effort. Due to the complexity of meshing codes, this is the only practical approach for keeping up with the ever-evolving sequential algorithms.		:# '''Code re-use''': a modular design of the parallel software that builds upon a previously designed sequential meshing code, such that it can be replaced and/or updated with a minimal effort. Due to the complexity of meshing codes, this is the only practical approach for keeping up with the ever-evolving sequential algorithms.
	:# '''Scalability''': the ratio of the time taken by the best sequential implementation to the time taken by the parallel implementation. The speedup is always limited by the inverse of the sequential fraction of the software, and therefore all non-trivial stages of the computation must be parallelized to leverage the current architectures with millions of cores.		:# '''Scalability''': the ratio of the time taken by the best sequential implementation to the time taken by the parallel implementation. The speedup is always limited by the inverse of the sequential fraction of the software, and therefore all non-trivial stages of the computation must be parallelized to leverage the current architectures with millions of cores.
			:# ''' Reproducibility ''': (weak & strong) TetGen meets none of these. For details see [LINK]. When it comes to shared memory no problem , in distributed memory (explain) This is why results are on limited geometries.
	The design and implementation of a sequential industrial strength code is labor intensive, it takes about 100 man-years. Parallel mesh generation code is even more labor intensive (by an order of magnitude for traditional parallel machines and expected to be higher for current and emerging architectures due to multiple memory and network hierarchies, fault-tolerance and power aware issues); however, because of the underlying theory that allows code re-use, this initial parallel implementation was achieved in less than six months with impressive functionality, i.e., the same as the sequential TetGen code.		The design and implementation of a sequential industrial strength code is labor intensive, it takes about 100 man-years. Parallel mesh generation code is even more labor intensive (by an order of magnitude for traditional parallel machines and expected to be higher for current and emerging architectures due to multiple memory and network hierarchies, fault-tolerance and power aware issues); however, because of the underlying theory that allows code re-use, this initial parallel implementation was achieved in less than six months with impressive functionality, i.e., the same as the sequential TetGen code.

Jbest: /* Using PDR (MPI Manual Page) */

2017-09-23T00:37:31Z

Using PDR (MPI Manual Page)

← Older revision		Revision as of 20:37, 22 September 2017
Line 238:		Line 238:

	<span style="color:red;"> Current alpha version of MPI PDR Release works only with the unit cube and bar3 geometries! </span>		<span style="color:red;"> Current alpha version of MPI PDR Release works only with the unit cube and bar3 geometries! </span>


	Running PDR without flags or arguments will display the following usage message:		Running PDR without flags or arguments will display the following usage message:

Jbest: /* Compiling PDR (Shared Memory Release) */

2017-09-23T00:35:40Z

Compiling PDR (Shared Memory Release)

← Older revision		Revision as of 20:35, 22 September 2017
Line 17:		Line 17:
	== Compiling PDR (Shared Memory Release) ==		== Compiling PDR (Shared Memory Release) ==
	The PDR source code is provided as a compressed tar archive, which is [http://www.cs.odu.edu/crtc/PDR/source/PDR_TetGen.tar.gz available here]. It can be extracted using the tar utility.		The PDR source code is provided as a compressed tar archive, which is [http://www.cs.odu.edu/crtc/PDR/source/PDR_TetGen.tar.gz available here]. It can be extracted using the tar utility.
	::: <span style="color:red;"> <syntaxhighlight lang="c"> tar xvzf PDR-buildnumber.tar.gz </syntaxhighlight> </span>		::: <span style="color:red;"> <syntaxhighlight lang="c"> tar xvzf PDR-buildnumber.tar.gz </syntaxhighlight> </span>

	After extracting PDR, cd into the PDR directory. To perform an out-of-source-build, create a compilation directory.		After extracting PDR, cd into the PDR directory. To perform an out-of-source-build, create a compilation directory.

Jbest: /* Using PDR (MPI Manual Page) */

2017-09-23T00:14:30Z

Using PDR (MPI Manual Page)

← Older revision		Revision as of 20:14, 22 September 2017
Line 243:		Line 243:
	: <span style="color:red;"> Usage: mpirun -np [numberofnodes] pdr -i [inputfilename] </span>		: <span style="color:red;"> Usage: mpirun -np [numberofnodes] pdr -i [inputfilename] </span>

	: <span style="color:red;"> -i -> Specify an input poly file (required) </span>		: <span style="color:red;"> -i -> Specify an input poly file (required) </span>
	: <span style="color:red;"> --grading -> Specify a uniform upper bound for circumradii </span>		: <span style="color:red;"> --grading -> Specify a uniform upper bound for circumradii </span>
	: <span style="color:red;"> --initial-depth -> Specify a depth limit for the initial octree </span>		: <span style="color:red;"> --initial-depth -> Specify a depth limit for the initial octree </span>
	</blockquote>		</blockquote>

Jbest: /* Running PDR (MPI Sample Execution) */

2017-09-23T00:08:26Z

Running PDR (MPI Sample Execution)

← Older revision		Revision as of 20:08, 22 September 2017
Line 270:		Line 270:
	=== Unit Cube ===		=== Unit Cube ===
	When run on the [http://www.cs.odu.edu/crtc/PDR/meshes/cube.poly.txt unit cube] using the following command:		When run on the [http://www.cs.odu.edu/crtc/PDR/meshes/cube.poly.txt unit cube] using the following command:
	::: <~~span style~~="~~color:red;~~"> mpirun -np 8 pdr --grading 0.015 --initial-depth 3 -i cube.poly </~~span~~>		::: <syntaxhighlight lang="c"> mpirun -np 8 pdr --grading 0.015 --initial-depth 3 -i cube.poly </syntaxhighlight>

	PDR will generate a mesh of approximately 2 million tetrahedra using eight nodes. The following shows the resulting mesh.		PDR will generate a mesh of approximately 2 million tetrahedra using eight nodes. The following shows the resulting mesh.
Line 284:		Line 284:
	[[File: Bar3.png\|200px]]		[[File: Bar3.png\|200px]]
	==== Command for 16-node Execution ====		==== Command for 16-node Execution ====
	::: <~~span style~~="~~color:red;~~"> mpirun -np 16 pdr --grading 0.015 --initial-depth 3 -i bar3.poly </~~span~~>		::: <syntaxhighlight lang="c"> mpirun -np 16 pdr --grading 0.015 --initial-depth 3 -i bar3.poly </syntaxhighlight>

	==== Argument Settings Used ====		==== Argument Settings Used ====
	<blockquote>		<blockquote>
	: <~~span style~~="~~color:red;~~"> --grading 0.015 </~~span~~> - set the grading		: <syntaxhighlight lang="c" inline> --grading 0.015 </syntaxhighlight> - set the grading
	: <~~span style~~="~~color:red;~~"> --initial-depth 3 </~~span~~> - set the intial octree depth		: <syntaxhighlight lang="c" inline> --initial-depth 3 </syntaxhighlight> - set the intial octree depth
	: <~~span style~~="~~color:red;~~"> -i bar3.poly </~~span~~> - set the input file		: <syntaxhighlight lang="c" inline> -i bar3.poly </syntaxhighlight> - set the input file
	</blockquote>		</blockquote>

Jbest: /* Input & Output */

2017-09-23T00:04:21Z

Input & Output

← Older revision		Revision as of 20:04, 22 September 2017
Line 250:		Line 250:
	=== Input & Output ===		=== Input & Output ===
	PDR requires that the –i flag be used to specify an input geometry.		PDR requires that the –i flag be used to specify an input geometry.
	::: <span style="color:red;"> mpirun -np 8 pdr –i cube.poly </span>		::: <span style="color:red;"> <syntaxhighlight lang="c"> mpirun -np 8 pdr –i cube.poly </syntaxhighlight> </span>

	By default, the base output filename will default to outputMesh.		By default, the base output filename will default to outputMesh.

Jbest: /* Running PDR (Shared Memory Sample Execution) */

2017-09-22T23:57:22Z

Running PDR (Shared Memory Sample Execution)

← Older revision		Revision as of 19:57, 22 September 2017
Line 127:		Line 127:

	==== Command for Single Core Execution ====		==== Command for Single Core Execution ====
	::: ~~<span style="color:red;">~~ <syntaxhighlight lang="c"> ./pdr -p 1 -s pdr-np-01-1.0-depth-04-06-brain_CBC3D.txt --grading 1.0 --initial-depth 4 -o pdr-np-01-1.0-depth-04-06-brain_CBC3D </syntaxhighlight~~> </span~~>		::: <syntaxhighlight lang="c"> ./pdr -p 1 -s pdr-np-01-1.0-depth-04-06-brain_CBC3D.txt --grading 1.0 --initial-depth 4 -o pdr-np-01-1.0-depth-04-06-brain_CBC3D </syntaxhighlight>

	==== Command for 32 Core Execution ====		==== Command for 32 Core Execution ====
	::: <~~span style~~="~~color:red;~~"> ./pdr -p 32 -s pdr-np-32-1.0-depth-04-06-brain_CBC3D.txt --grading 1.0 --initial-depth 4 --max-depth 6 -o pdr-np-32-1.0-depth-04-06-brain_CBC3D </~~span~~>		::: <syntaxhighlight lang="c"> ./pdr -p 32 -s pdr-np-32-1.0-depth-04-06-brain_CBC3D.txt --grading 1.0 --initial-depth 4 --max-depth 6 -o pdr-np-32-1.0-depth-04-06-brain_CBC3D </syntaxhighlight>

	==== Argument Settings Used ====		==== Argument Settings Used ====
	<blockquote>		<blockquote>
	: <~~span style~~="~~color:red;~~"> -p 1 </~~span~~> - set the number of cores		: <syntaxhighlight lang="c" inline> -p 1 </syntaxhighlight> - set the number of cores
	: <~~span style~~="~~color:red;~~"> -s pdr-np-01-1.0-depth-04-06-brain_CBC3D.txt </~~span~~> - set the file name for time statistic output		: <syntaxhighlight lang="c" inline> -s pdr-np-01-1.0-depth-04-06-brain_CBC3D.txt </syntaxhighlight> - set the file name for time statistic output
	: <~~span style~~="~~color:red;~~"> --grading 1.0 </~~span~~> - set the grading		: <syntaxhighlight lang="c" inline> --grading 1.0 </syntaxhighlight> - set the grading
	: <~~span style~~="~~color:red;~~"> --initial-depth 4 </~~span~~> - set the initial octree depth		: <syntaxhighlight lang="c" inline> --initial-depth 4 </syntaxhighlight> - set the initial octree depth
	: <~~span style~~="~~color:red;~~"> --max-depth 6 </~~span~~> - set the maximum octree depth (ignored in non-progressive mode)		: <syntaxhighlight lang="c" inline> --max-depth 6 </syntaxhighlight> - set the maximum octree depth (ignored in non-progressive mode)
	: <~~span style~~="~~color:red;~~"> -o pdr-np-01-1.0-depth-04-06-brain_CBC3D </~~span~~> - set the base output file name		: <syntaxhighlight lang="c" inline> -o pdr-np-01-1.0-depth-04-06-brain_CBC3D </syntaxhighlight> - set the base output file name
	: <~~span style~~="~~color:red;~~"> -i brain_CBC3D.poly </~~span~~> - set the input file		: <syntaxhighlight lang="c" inline> -i brain_CBC3D.poly </syntaxhighlight> - set the input file
	</blockquote>		</blockquote>

Line 165:		Line 165:

	==== Command for Single Core Execution ====		==== Command for Single Core Execution ====
	::: <~~span style~~="~~color:red;~~"> ./pdr -p 1 -s pdr-np-01-1.0-depth-04-06-socket.txt --grading 1.0 --initial-depth 4 --max-depth 6 -o pdr-np-01-1.0-depth-04-06-socket -i socket.poly </~~span~~>		::: <syntaxhighlight lang="c"> ./pdr -p 1 -s pdr-np-01-1.0-depth-04-06-socket.txt --grading 1.0 --initial-depth 4 --max-depth 6 -o pdr-np-01-1.0-depth-04-06-socket -i socket.poly </syntaxhighlight>

	==== Command for 32 Core Execution ====		==== Command for 32 Core Execution ====
	::: <~~span style~~="~~color:red;~~"> ./pdr -p 32 -s pdr-np-32-1.0-depth-04-06-socket.txt --grading 1.0 --initial-depth 4 --max-depth 6 -o pdr-np-32-1.0-depth-04-06-socket -i socket.poly </~~span~~>		::: <syntaxhighlight lang="c"> ./pdr -p 32 -s pdr-np-32-1.0-depth-04-06-socket.txt --grading 1.0 --initial-depth 4 --max-depth 6 -o pdr-np-32-1.0-depth-04-06-socket -i socket.poly </syntaxhighlight>

	==== Argument Settings Used ====		==== Argument Settings Used ====
	<blockquote>		<blockquote>
	: <~~span style~~="~~color:red;~~"> -p 1 </~~span~~> - set the number of cores		: <syntaxhighlight lang="c" inline> -p 1 </syntaxhighlight> - set the number of cores
	: <~~span style~~="~~color:red;~~"> -s pdr-np-01-1.0-depth-04-06-socket.txt </~~span~~> - set the file name for time statistic output		: <syntaxhighlight lang="c" inline> -s pdr-np-01-1.0-depth-04-06-socket.txt </syntaxhighlight> - set the file name for time statistic output
	: <~~span style~~="~~color:red;~~"> --grading 1.0 </~~span~~> - set the grading		: <syntaxhighlight lang="c" inline> --grading 1.0 </syntaxhighlight> - set the grading
	: <~~span style~~="~~color:red;~~"> --initial-depth 4 </~~span~~> - set the initial octree depth		: <syntaxhighlight lang="c" inline> --initial-depth 4 </syntaxhighlight> - set the initial octree depth
	: <~~span style~~="~~color:red;~~"> --max-depth 6 </~~span~~> - set the maximum octree depth (ignored in non-progressive mode)		: <syntaxhighlight lang="c" inline> --max-depth 6 </syntaxhighlight> - set the maximum octree depth (ignored in non-progressive mode)
	: <~~span style~~="~~color:red;~~"> -o pdr-np-01-1.0-depth-04-06-socket </~~span~~> - set the base output file name		: <syntaxhighlight lang="c" inline> -o pdr-np-01-1.0-depth-04-06-socket </syntaxhighlight> - set the base output file name
	: <~~span style~~="~~color:red;~~"> -i socket.poly </~~span~~> - set the input file		: <syntaxhighlight lang="c" inline> -i socket.poly </syntaxhighlight> - set the input file
	</blockquote>		</blockquote>

Line 204:		Line 204:

	==== Command for Single Core Execution ====		==== Command for Single Core Execution ====
	::: <~~span style~~="~~color:red;~~"> ./pdr -p 1 -s pdr-np-01-0.03-depth-04-06-bar3.txt --grading 0.03 --initial-depth 4 --max-depth 6 -o pdr-np-01-0.03-depth-04-06-bar3 -i bar3.poly </~~span~~>		::: <syntaxhighlight lang="c"> ./pdr -p 1 -s pdr-np-01-0.03-depth-04-06-bar3.txt --grading 0.03 --initial-depth 4 --max-depth 6 -o pdr-np-01-0.03-depth-04-06-bar3 -i bar3.poly </syntaxhighlight>

	==== Command for 32 Core Execution ====		==== Command for 32 Core Execution ====
	::: <~~span style~~="~~color:red;~~"> ./pdr -p 32 -s pdr-np-32-0.03-depth-04-06-bar3.txt --grading 0.03 --initial-depth 4 --max-depth 6 -o pdr-np-32-0.03-depth-04-06-bar3 -i bar3.poly </~~span~~>		::: <syntaxhighlight lang="c"> ./pdr -p 32 -s pdr-np-32-0.03-depth-04-06-bar3.txt --grading 0.03 --initial-depth 4 --max-depth 6 -o pdr-np-32-0.03-depth-04-06-bar3 -i bar3.poly </syntaxhighlight>

	==== Argument Settings Used ====		==== Argument Settings Used ====
	<blockquote>		<blockquote>
	: <~~span style~~="~~color:red;~~"> -p 1 </~~span~~> - set the number of cores		: <syntaxhighlight lang="c" inline> -p 1 </syntaxhighlight> - set the number of cores
	: <~~span style~~="~~color:red;~~"> -s pdr-np-01-0.03-depth-04-06-bar3.txt </~~span~~> - set the file name for time statistic output		: <syntaxhighlight lang="c" inline> -s pdr-np-01-0.03-depth-04-06-bar3.txt </syntaxhighlight> - set the file name for time statistic output
	: <~~span style~~="~~color:red;~~"> --grading 0.03 </~~span~~> - set the grading		: <syntaxhighlight lang="c" inline> --grading 0.03 </syntaxhighlight> - set the grading
	: <~~span style~~="~~color:red;~~"> --initial-depth 4 </~~span~~> - set the initial octree depth		: <syntaxhighlight lang="c" inline> --initial-depth 4 </syntaxhighlight> - set the initial octree depth
	: <~~span style~~="~~color:red;~~"> --max-depth 6 </~~span~~> - set the maximum octree depth (ignored in non-progressive mode)		: <syntaxhighlight lang="c" inline> --max-depth 6 </syntaxhighlight> - set the maximum octree depth (ignored in non-progressive mode)
	: <~~span style~~="~~color:red;~~"> -o pdr-np-01-0.03-depth-04-06-bar3 </~~span~~> - set the base output file name		: <syntaxhighlight lang="c" inline> -o pdr-np-01-0.03-depth-04-06-bar3 </syntaxhighlight> - set the base output file name
	: <~~span style~~="~~color:red;~~"> -i bar3.poly </~~span~~> - set the input file		: <syntaxhighlight lang="c" inline> -i bar3.poly </syntaxhighlight> - set the input file
	</blockquote>		</blockquote>

Jbest: /* Command for Single Core Execution */

2017-09-22T23:36:41Z

Command for Single Core Execution

← Older revision		Revision as of 19:36, 22 September 2017
Line 127:		Line 127:

	==== Command for Single Core Execution ====		==== Command for Single Core Execution ====
	::: <span style="color:red;"> ./pdr -p 1 -s pdr-np-01-1.0-depth-04-06-brain_CBC3D.txt --grading 1.0 --initial-depth 4 -o pdr-np-01-1.0-depth-04-06-brain_CBC3D </span>		::: <span style="color:red;"> <syntaxhighlight lang="c"> ./pdr -p 1 -s pdr-np-01-1.0-depth-04-06-brain_CBC3D.txt --grading 1.0 --initial-depth 4 -o pdr-np-01-1.0-depth-04-06-brain_CBC3D </syntaxhighlight> </span>

	==== Command for 32 Core Execution ====		==== Command for 32 Core Execution ====