Skip to content

Instantly share code, notes, and snippets.

View tkphd's full-sized avatar
🏳️‍🌈

Trevor Keller tkphd

🏳️‍🌈
74 followers · 41 following
View GitHub Profile
@tkphd
tkphd / mpitest.py
Created August 25, 2015 16:46
# Based on https://gist.github.com/wd15/871797
import time
import fipy as fp
import numpy as np
from mpi4py import MPI
np.random.seed(1)
L = 1.
N = 30
@tkphd
tkphd / setvalue_example.py
Created September 9, 2015 14:38
from fipy import *
from scipy import interpolate
import numpy as np
mesh = Grid1D(dx=3, nx=10)
datx = mesh.cellCenters[0]
daty = np.sin(4*np.pi*datx/10)
valumesh = Grid1D(dx=1, nx=10)
@tkphd
tkphd / cartesian_grid.py
Created September 17, 2015 22:15
Script to test uniform grid memory access time
from fipy import *
from scipy import interpolate
import numpy as np
##Create mesh in the Cartesian Coordinate system (X,Y)
nx = 100
ny = nx
dx = 1.0
dy = dx
CL = nx*dx
@tkphd
tkphd / polar_grid.py
Created September 17, 2015 22:17
Script to test polar grid memory access time
from fipy import *
from scipy import interpolate
import numpy as np
##Create mesh in Polar Coordinate system (LnR,Cta)
nLnR = 100
nCta = 50
dLnR = 0.02
dCta = 2*np.pi/nCta
PL = nLnR*dLnR
@tkphd
tkphd / Dockerfile
Last active May 13, 2016 21:31
Docker build for MMSP spinodal decomposition
# Docker file to install MMSP in a Ubuntu container
# with a Jupyter notebook exposed on port 8888
#
# Cloned from https://gist.github.com/wd15/d233c706bb493c91ecd5
#
FROM ubuntu:15.10
MAINTAINER Trevor Keller "https://github.com/tkphd"
## Install only required packages for making a user as the
@tkphd
tkphd / meta.yml
Last active January 5, 2017 00:42
MMSP spinodal decomposition benchmark, periodic domain
---
metadata:
# Describe the runtime environment
summary: MPI parallel workstation benchmark with MMSP, periodic domain
author: Trevor Keller
email: [email protected]
date: Wed, 04 Jan 2017 19:30:31 -0500
hardware:
# Required hardware details
architecture: x86_64
@tkphd
tkphd / mmsp_1a_raspi_results.csv
Last active February 10, 2017 17:27
MMSP spinodal decomposition benchmark on Raspberry Pi 3 cluster (4 nodes)
time free_energy
0 319.094
41.6667 175.655
83.3333 140.518
125 128.394
166.667 121.085
208.333 115.449
250 110.879
291.667 107.212
333.333 104.281
@tkphd
tkphd / mmsp_1a_wrksttn.csv
Last active February 15, 2017 16:32
MMSP workstation spinodal decomposition results
time free_energy
0 319.094
106.667 132.607
213.333 114.85
320 105.152
426.667 99.4598
533.333 95.5889
640 92.5611
746.667 89.9498
853.333 87.5588
@tkphd
tkphd / CTCMS_sbatch.sh
Created March 16, 2017 19:01
#!/bin/bash
#SBATCH --partition rack6i # -p, partition
#SBATCH --time 12:00:00 # -t, time (hh:mm:ss or dd-hh:mm:ss)
#SBATCH --nodes=2 # total number of machines
#SBATCH --ntasks=4 # 2 sockets per rack6i machine
#SBATCH --cpus-per-task=28 # 28 OMP threads per rack6i MPI rank
#SBATCH -J Alloy625
#SBATCH -D /data/tnk10/phase-field/alloy625/run2/TKR4p074
export OMP_NUM_THREADS=$SLURM_CPUS_PER_TASK
@tkphd
tkphd / Du_Cr-Nb-Ni_simple.tdb
Created March 29, 2017 15:00
$ Database file written 1900- 1-**
$ From database: USER
ELEMENT /- ELECTRON_GAS 0.0000E+00 0.0000E+00 0.0000E+00!
ELEMENT VA VACUUM 0.0000E+00 0.0000E+00 0.0000E+00!
ELEMENT CR BCC_A2 5.1996E+01 4.0500E+03 2.3560E+01!
ELEMENT NB BCC_A2 9.2906E+01 5.2200E+03 3.6270E+01!
ELEMENT NI FCC_A1 5.8690E+01 4.7870E+03 2.9796E+01!
FUNCTION GLIQCR 298.15 +24339.955-11.420225*T+2.37615E-21*T**7+GHSERCR#;