Example: float_32_64_benchmark

Note

You can launch an interactive, editable version of this example without installing any local files using the Binder service (although note that at some times this may be slow or fail to open): launchbinder

Benchmark showing the performance of float32 versus float64.

from brian2 import *
from brian2.devices.device import reset_device, reinit_devices

# CUBA benchmark
def run_benchmark(name):
    if name=='CUBA':

        taum = 20*ms
        taue = 5*ms
        taui = 10*ms
        Vt = -50*mV
        Vr = -60*mV
        El = -49*mV

        eqs = '''
        dv/dt  = (ge+gi-(v-El))/taum : volt (unless refractory)
        dge/dt = -ge/taue : volt
        dgi/dt = -gi/taui : volt
        '''

        P = NeuronGroup(4000, eqs, threshold='v>Vt', reset='v = Vr', refractory=5*ms,
                        method='exact')
        P.v = 'Vr + rand() * (Vt - Vr)'
        P.ge = 0*mV
        P.gi = 0*mV

        we = (60*0.27/10)*mV # excitatory synaptic weight (voltage)
        wi = (-20*4.5/10)*mV # inhibitory synaptic weight
        Ce = Synapses(P, P, on_pre='ge += we')
        Ci = Synapses(P, P, on_pre='gi += wi')
        Ce.connect('i<3200', p=0.02)
        Ci.connect('i>=3200', p=0.02)

    elif name=='COBA':

        # Parameters
        area = 20000 * umetre ** 2
        Cm = (1 * ufarad * cm ** -2) * area
        gl = (5e-5 * siemens * cm ** -2) * area

        El = -60 * mV
        EK = -90 * mV
        ENa = 50 * mV
        g_na = (100 * msiemens * cm ** -2) * area
        g_kd = (30 * msiemens * cm ** -2) * area
        VT = -63 * mV
        # Time constants
        taue = 5 * ms
        taui = 10 * ms
        # Reversal potentials
        Ee = 0 * mV
        Ei = -80 * mV
        we = 6 * nS  # excitatory synaptic weight
        wi = 67 * nS  # inhibitory synaptic weight

        # The model
        eqs = Equations('''
        dv/dt = (gl*(El-v)+ge*(Ee-v)+gi*(Ei-v)-
                 g_na*(m*m*m)*h*(v-ENa)-
                 g_kd*(n*n*n*n)*(v-EK))/Cm : volt
        dm/dt = alpha_m*(1-m)-beta_m*m : 1
        dn/dt = alpha_n*(1-n)-beta_n*n : 1
        dh/dt = alpha_h*(1-h)-beta_h*h : 1
        dge/dt = -ge*(1./taue) : siemens
        dgi/dt = -gi*(1./taui) : siemens
        alpha_m = 0.32*(mV**-1)*4*mV/exprel((13*mV-v+VT)/(4*mV))/ms : Hz
        beta_m = 0.28*(mV**-1)*5*mV/exprel((v-VT-40*mV)/(5*mV))/ms : Hz
        alpha_h = 0.128*exp((17*mV-v+VT)/(18*mV))/ms : Hz
        beta_h = 4./(1+exp((40*mV-v+VT)/(5*mV)))/ms : Hz
        alpha_n = 0.032*(mV**-1)*5*mV/exprel((15*mV-v+VT)/(5*mV))/ms : Hz
        beta_n = .5*exp((10*mV-v+VT)/(40*mV))/ms : Hz
        ''')

        P = NeuronGroup(4000, model=eqs, threshold='v>-20*mV', refractory=3 * ms,
                        method='exponential_euler')
        Pe = P[:3200]
        Pi = P[3200:]
        Ce = Synapses(Pe, P, on_pre='ge+=we')
        Ci = Synapses(Pi, P, on_pre='gi+=wi')
        Ce.connect(p=0.02)
        Ci.connect(p=0.02)

        # Initialization
        P.v = 'El + (randn() * 5 - 5)*mV'
        P.ge = '(randn() * 1.5 + 4) * 10.*nS'
        P.gi = '(randn() * 12 + 20) * 10.*nS'

    run(1 * second, profile=True)

    return sum(t for name, t in magic_network.profiling_info)

def generate_results(num_repeats):
    results = {}

    for name in ['CUBA', 'COBA']:
        for target in ['numpy', 'cython']:
            for dtype in [float32, float64]:
                prefs.codegen.target = target
                prefs.core.default_float_dtype = dtype
                times = [run_benchmark(name) for repeat in range(num_repeats)]
                results[name, target, dtype.__name__] = amin(times)

    for name in ['CUBA', 'COBA']:
        for dtype in [float32, float64]:
            times = []
            for _ in range(num_repeats):
                reset_device()
                reinit_devices()
                set_device('cpp_standalone', directory=None, with_output=False)
                prefs.core.default_float_dtype = dtype
                times.append(run_benchmark(name))
            results[name, 'cpp_standalone', dtype.__name__] = amin(times)

    return results

results = generate_results(3)

bar_width = 0.9
names = ['CUBA', 'COBA']
targets = ['numpy', 'cython', 'cpp_standalone']
precisions = ['float32', 'float64']

figure(figsize=(8, 8))
for j, name in enumerate(names):
    subplot(2, 2, 1+2*j)
    title(name)
    index = arange(len(targets))
    for i, precision in enumerate(precisions):
        bar(index+i*bar_width/len(precisions),
            [results[name, target, precision] for target in targets],
            bar_width/len(precisions), label=precision, align='edge')
    ylabel('Time (s)')
    if j:
        xticks(index+0.5*bar_width, targets, rotation=45)
    else:
        xticks(index+0.5*bar_width, ('',)*len(targets))
        legend(loc='best')

    subplot(2, 2, 2+2*j)
    index = arange(len(precisions))
    for i, target in enumerate(targets):
        bar(index+i*bar_width/len(targets),
            [results[name, target, precision] for precision in precisions],
            bar_width/len(targets), label=target, align='edge')
    ylabel('Time (s)')
    if j:
        xticks(index+0.5*bar_width, precisions, rotation=45)
    else:
        xticks(index+0.5*bar_width, ('',)*len(precisions))
        legend(loc='best')

tight_layout()
show()