Cell splitting in neural networks extends strong scaling (Hines et al. 2008)

 Download zip file   Auto-launch 
Help downloading and running models
Accession:97917
Neuron tree topology equations can be split into two subtrees and solved on different processors with no change in accuracy, stability, or computational effort; communication costs involve only sending and receiving two double precision values by each subtree at each time step. Application of the cell splitting method to two published network models exhibits good runtime scaling on twice as many processors as could be effectively used with whole-cell balancing.
Reference:
1 . Hines ML, Eichner H, Schürmann F (2008) Neuron splitting in compute-bound parallel network simulations enables runtime scaling with twice as many processors. J Comput Neurosci 25:203-10 [PubMed]
Citations  Citation Browser
Model Information (Click on a link to find other models with that property)
Model Type: Realistic Network;
Brain Region(s)/Organism: Generic;
Cell Type(s):
Channel(s):
Gap Junctions:
Receptor(s):
Gene(s):
Transmitter(s):
Simulation Environment: NEURON;
Model Concept(s): Methods;
Implementer(s): Hines, Michael [Michael.Hines at Yale.edu];
/
splitcell
pardentategyrus
readme.html *
bgka.mod *
CaBK.mod *
ccanl.mod *
Gfluct2.mod *
gskch.mod *
hyperde3.mod *
ichan2.mod *
LcaMig.mod *
nca.mod *
tca.mod *
bg.sh
DG500_M7.hoc *
dgnetactivity.jpg *
dgnettraces.jpg *
init.hoc
initorig.hoc *
modstat *
mosinit_orig.hoc *
out.std
parRI10sp.hoc
RI10sp.hoc
test1.sh *
time *
                            
TITLE nca.mod  
 
COMMENT
konduktivitas valtozas hatasa- somaban 
ENDCOMMENT
 
UNITS {
        (mA) =(milliamp)
        (mV) =(millivolt)
        (uF) = (microfarad)
	(molar) = (1/liter)
	(nA) = (nanoamp)
	(mM) = (millimolar)
	(um) = (micron)
	FARADAY = 96520 (coul)
	R = 8.3134	(joule/degC)
}
 
? interface 
NEURON { 
SUFFIX nca
USEION nca READ enca WRITE inca VALENCE 2 
RANGE  gnca
RANGE gncabar
RANGE cinf, ctau, dinf, dtau, inca
}
 
INDEPENDENT {t FROM 0 TO 100 WITH 100 (ms)}
 
PARAMETER {
        v (mV) 
        celsius = 6.3 (degC)
        dt (ms) 
	gncabar (mho/cm2)
}
 
STATE {
	c d
}
 
ASSIGNED {
	  gnca (mho/cm2)
	inca (mA/cm2)
	enca (mV)

	cinf dinf
	ctau (ms) dtau (ms) 
	cexp dexp      
} 

? currents
BREAKPOINT {
	SOLVE states
        gnca = gncabar*c*c*d
	inca = gnca*(v-enca)
}
 
UNITSOFF
 
INITIAL {
	trates(v)
	c = cinf
	d = dinf
}

? states
PROCEDURE states() {	:Computes state variables m, h, and n 
        trates(v)	:      at the current v and dt.
	c = c + cexp*(cinf-c)
	d = d + dexp*(dinf-d)
        VERBATIM
        return 0;
        ENDVERBATIM
}
 
LOCAL q10

? rates
PROCEDURE rates(v) {  :Computes rate and other constants at current v.
                      :Call once from HOC to initialize inf at resting v.
        LOCAL  alpha, beta, sum
       q10 = 3^((celsius - 6.3)/10)
                :"c" NCa activation system
        alpha = -0.19*vtrap(v-19.88,-10)
	beta = 0.046*exp(-v/20.73)
	sum = alpha+beta        
	ctau = 1/sum      cinf = alpha/sum
                :"d" NCa inactivation system
	alpha = 0.00016/exp(-v/48.4)
	beta = 1/(exp((-v+39)/10)+1)
	sum = alpha+beta        
	dtau = 1/sum      dinf = alpha/sum
}
 
PROCEDURE trates(v) {  :Computes rate and other constants at current v.
                      :Call once from HOC to initialize inf at resting v.
	LOCAL tinc
        TABLE  cinf, cexp, dinf, dexp, ctau, dtau
	DEPEND dt, celsius FROM -100 TO 100 WITH 200
                           
	rates(v)	: not consistently executed from here if usetable_hh == 1
		: so don't expect the tau values to be tracking along with
		: the inf values in hoc

	       tinc = -dt * q10
	cexp = 1 - exp(tinc/ctau)
	dexp = 1 - exp(tinc/dtau)
}
 
FUNCTION vtrap(x,y) {  :Traps for 0 in denominator of rate eqns.
        if (fabs(x/y) < 1e-6) {
                vtrap = y*(1 - x/y/2)
        }else{  
                vtrap = x/(exp(x/y) - 1)
        }
}
 
UNITSON