The PyCUDA module
Day: Monday, December 2Time: 2:10 PM - 3:00 PM
Place: Snedecor Hall 2113
PyCUDA is a GPU-enabled Python module that weaves Python with CUDA C. PyCUDA allows the user to build and execute GPU kernels within Python, and it encapsulates the transfer of Python data objects to and from the GPU. Hence, PyCUDA allows the user to take advantage of both the computing power of CUDA C and the convenience of Python. This talk will introduce PyCUDA with several example scripts and conclude with a brief overview of ABC-SysBio, a GPU-accelerated approximate Bayesian computation utility for estimating parameters in dynamical systems.
Materials
SlidesVideo
Python Code
demo.pydemohandler.py
DemoMetaCodepy.py
demoshort.py
DumpProperties.py
functiontemplates.py
hello_gpu.py
MatmulSimple.py
MeasureGpuarraySpeedRandom.py
pycurand.py
reduction.py
scan.py
ABC-SysBio Examples
Example 1
data.txtinput_file_SIR_auto.xml
input_file_SIR_normk.xml
input_file_SIR_simulation.xml
input_file_SIR_x0_normk.xml
input_file_SIR_x0.xml
input_file_SIR.xml
SIRModel1.xml
SIRModel2.xml
SIRModel3.xml
Example 3
DimerisationDiscreetStochastic.xmlinput_file_dimerisation_auto.xml
input_file_dimerisation.xml
Example 5
input_file_repressilator_auto.xmlinput_file_repressilator.xml
repressilator.xml
Example 2
input_file_lingrow_normk.xmlinput_file_lingrow_normk2.xml
input_file_lingrow_restart.xml.xml
input_file_lingrow_simulation.xml
input_file_lingrow_x0_restart.xml
input_file_lingrow_x0.xml
input_file_lingrow.xml
LinearGrowthModelModified.py
Example 4
immigration-death.xmlinput_file_pop_mjp.xml
input_file_pop_ode.xml
input_file_pop_sde.xml
logistic.xml
Example 6
customABC.pyinput_file_adaptation.xml
three_nodes_model1.xml
three_nodes_model2.xml
three_nodes_model3.xml
three_nodes_model4.xml
three_nodes_model5.xml
three_nodes_model6.xml
three_nodes_model7.xml
three_nodes_model8.xml
three_nodes_model9.xml
three_nodes_model10.xml
three_nodes_model11.xml