Computational XRD - Revision history

Alank2: /* Making Nanoparticles */

2015-08-23T21:57:52Z

Making Nanoparticles

← Older revision		Revision as of 21:57, 23 August 2015
Line 13:		Line 13:

	and these can be run with the command:		and these can be run with the command:
			<pre>
	bin/eam generate_particles.py		bin/eam generate_particles.py
			</pre>

	'''Notable Parameters'''		'''Notable Parameters'''

Alank2: /* Intensity Calculation in C++ */

2015-08-23T21:57:22Z

Intensity Calculation in C++

← Older revision		Revision as of 21:57, 23 August 2015
Line 30:		Line 30:
	===Intensity Calculation in C++===		===Intensity Calculation in C++===

	First, make sure the required libraries are installed. The GSL and C++11 packages are required (the GSL library is for easily making histograms). The required files are the previous file of atomic positions and the ~~attached~~ C++ file:		First, make sure the required libraries are installed. The GSL and C++11 packages are required (the GSL library is for easily making histograms). The required files are the previous file of atomic positions written by generate_particles.py and the following C++ file which can be run as:

	Now we are ready to run

	<pre>		<pre>

Alank2: /* Making Nanoparticles */

2015-08-23T21:56:37Z

Making Nanoparticles

← Older revision		Revision as of 21:56, 23 August 2015
Line 9:		Line 9:

	Here we give an example of how to generate nanoparticles with different dislocations. One may have to play around a bit with the directories to make sure that the files get opened and written to the correct places. A python version of the MD++ code is necessary. The relevant python files are:		Here we give an example of how to generate nanoparticles with different dislocations. One may have to play around a bit with the directories to make sure that the files get opened and written to the correct places. A python version of the MD++ code is necessary. The relevant python files are:
	~~[[media:~~generate_particles.py ~~\| generate_particles.py]]~~		generate_particles.py

	XRDsetup.py		XRDsetup.py

Alank2: /* Making Nanoparticles */

2015-08-23T21:54:23Z

Making Nanoparticles

← Older revision		Revision as of 21:54, 23 August 2015
Line 9:		Line 9:

	Here we give an example of how to generate nanoparticles with different dislocations. One may have to play around a bit with the directories to make sure that the files get opened and written to the correct places. A python version of the MD++ code is necessary. The relevant python files are:		Here we give an example of how to generate nanoparticles with different dislocations. One may have to play around a bit with the directories to make sure that the files get opened and written to the correct places. A python version of the MD++ code is necessary. The relevant python files are:
			[[media:generate_particles.py \| generate_particles.py]]

	generate_particles.py
	XRDsetup.py		XRDsetup.py

Line 27:		Line 27:
	</pre>		</pre>

	This generates a series of .cn, .cfg, and a file of atomic positions which will be read by the C++ code to calculate the Intensity.		This generates a series of .cn, .cfg, and a file of atomic positions which will be read by the C++ code to calculate the Intensity.

	===Intensity Calculation in C++===		===Intensity Calculation in C++===

Alank2: /* CUDA Code */

2015-08-23T21:50:55Z

CUDA Code

← Older revision		Revision as of 21:50, 23 August 2015
Line 83:		Line 83:

	===CUDA Code===		===CUDA Code===
	The code is obviously parallel. The next step is to re-write the code in CUDA to improve speed. The vast majority of time is spent in the distances calculation so for now, only this section will be made into a CUDA kernel. As the particles get bigger and more heavily deformed, the number of bins may need to be increased, leading to a notable increase in the XRD time. Unless the number of bins necessary is outrageously large (even at the huge overkill of 100,000 bins for a 4 nm nanoparticle, the entire calculation of distances + XRD takes ), the XRD time should not get too large from this contribution. One important note is that the cluster of GPUs need to have the ~~gsl~~ libraries available.		The code is obviously parallel. The next step is to re-write the code in CUDA to improve speed. The vast majority of time is spent in the distances calculation so for now, only this section will be made into a CUDA kernel. As the particles get bigger and more heavily deformed, the number of bins may need to be increased, leading to a notable increase in the XRD time. Unless the number of bins necessary is outrageously large (even at the huge overkill of 100,000 bins for a 4 nm nanoparticle, the entire calculation of distances + XRD takes 75 seconds), the XRD time should not get too large from this contribution. One important note is that the cluster of GPUs need to have the GSL libraries available.

Alank2: /* CUDA Code */

2015-08-23T21:47:51Z

CUDA Code

← Older revision		Revision as of 21:47, 23 August 2015
Line 83:		Line 83:

	===CUDA Code===		===CUDA Code===
	The code is obviously parallel. The next step is to re-write the code in CUDA to improve speed. The vast majority of time is spent in the distances calculation so for now, only this section will be made into a CUDA kernel. One important note is that the cluster of GPUs need to have the gsl libraries available.		The code is obviously parallel. The next step is to re-write the code in CUDA to improve speed. The vast majority of time is spent in the distances calculation so for now, only this section will be made into a CUDA kernel. As the particles get bigger and more heavily deformed, the number of bins may need to be increased, leading to a notable increase in the XRD time. Unless the number of bins necessary is outrageously large (even at the huge overkill of 100,000 bins for a 4 nm nanoparticle, the entire calculation of distances + XRD takes ), the XRD time should not get too large from this contribution. One important note is that the cluster of GPUs need to have the gsl libraries available.

Alank2: /* Intensity Calculation in C++ */

2015-08-23T21:43:19Z

Intensity Calculation in C++

← Older revision		Revision as of 21:43, 23 August 2015
Line 31:		Line 31:
	===Intensity Calculation in C++===		===Intensity Calculation in C++===

	First, make sure the required libraries are installed. The GSL, C++11~~, and OpenMP~~ packages are required (the GSL library is for easily making histograms). The required files are the previous file of atomic positions and the attached C++ file:		First, make sure the required libraries are installed. The GSL and C++11 packages are required (the GSL library is for easily making histograms). The required files are the previous file of atomic positions and the attached C++ file:

	Now we are ready to run		Now we are ready to run

	<pre>		<pre>
	g++ -o XRD.exe ~~XRD_OpenMP~~.cpp -lgsl -lgslcblas -lm ~~-fopenmp~~ -std=c++11		g++ -o XRD.exe XRD.cpp -lgsl -lgslcblas -lm -std=c++11
	</pre>		</pre>
	followed simply by		followed simply by
Line 51:		Line 51:

	wavelength = 0.67 # the wavelength of the incident monochromatic X-ray beam		wavelength = 0.67 # the wavelength of the incident monochromatic X-ray beam

	num_bins = 25000 # the number of bins in the histogram per division		num_bins = 25000 # the number of bins in the histogram per division
	N = ~~10000~~ # ~~the~~ ~~number~~ of ~~points~~ ~~for~~ the ~~Intensity~~ ~~and~~ Q		# the background noise is quite dependent on this
			# parameter so don't make it too small

			N = 10000 # the number of Intensity points to calculate
	</pre>		</pre>

	I plan to run several different sizes and compare the time taken for this calculation for a fixed num_bins, divs, and N to access the scalability. Currently the code is highly un-optimized and still runs very fast.		I plan to run several different sizes and compare the time taken for this calculation for a fixed num_bins, divs, and N to access the scalability. Currently the code is highly un-optimized (completely serial) and still runs very fast. A 4 nm nanoparticle calculation takes ~30 seconds with this code.

	===Plotting and Interpolating===		===Plotting and Interpolating===

Alank2: /* Intensity Calculation in C++ */

2015-08-23T21:39:20Z

Intensity Calculation in C++

← Older revision		Revision as of 21:39, 23 August 2015
Line 51:		Line 51:

	wavelength = 0.67 # the wavelength of the incident monochromatic X-ray beam		wavelength = 0.67 # the wavelength of the incident monochromatic X-ray beam
	~~divs~~ = 6 # the number of ~~partial~~ ~~distance~~ ~~calculations~~ to ~~compute.~~		num_bins = 25000 # the number of bins in the histogram per division
	# ~~this~~ ~~needs~~ to be ~~adjusted~~ to ~~larger values for~~ ~~large~~ ~~runs~~		N = 10000 # the number of points for the Intensity and Q
	# because of memory issues. If not set properly, the code will
	# exit with a memory error.
	num_bins = 5000 # the number of bins in the histogram per division
	N = 1000 # the number of points for the Intensity and Q
	</pre>		</pre>

	I plan to run several different sizes and compare the time taken for this calculation for a fixed num_bins, divs, and N to access the scalability. Currently the code is highly un-optimized. ~~This~~ ~~code~~ ~~should~~ be ~~re-written with dynamic arrays instead of vectors in order to become much more parallel~~.		I plan to run several different sizes and compare the time taken for this calculation for a fixed num_bins, divs, and N to access the scalability. Currently the code is highly un-optimized and still runs very fast.

	===Plotting and Interpolating===		===Plotting and Interpolating===

Alank2: /* CUDA Code */

2015-08-23T20:35:40Z

CUDA Code

← Older revision		Revision as of 20:35, 23 August 2015
Line 83:		Line 83:

	===CUDA Code===		===CUDA Code===
	The code is obviously parallel. The next step is to re-write the code in CUDA to improve speed. The vast majority of time is spent in the distances calculation so for now, only this section will be made into a CUDA kernel.		The code is obviously parallel. The next step is to re-write the code in CUDA to improve speed. The vast majority of time is spent in the distances calculation so for now, only this section will be made into a CUDA kernel. One important note is that the cluster of GPUs need to have the gsl libraries available.

Alank2: /* CUDA Code */

2015-08-23T19:03:10Z

CUDA Code

← Older revision		Revision as of 19:03, 23 August 2015
Line 83:		Line 83:

	===CUDA Code===		===CUDA Code===
	The code is obviously parallel. The next step is to re-write the code in CUDA to improve speed.		The code is obviously parallel. The next step is to re-write the code in CUDA to improve speed. The vast majority of time is spent in the distances calculation so for now, only this section will be made into a CUDA kernel.