GeNN  3.3.0
GPU enhanced Neuronal Networks (GeNN)
global.h File Reference

Global header file containing a few global variables. Part of the code generation section. More...

#include <string>
#include <cuda.h>
#include <cuda_runtime.h>
#include "variableMode.h"

Go to the source code of this file.

Namespaces

 GENN_FLAGS
 
 GENN_PREFERENCES
 

Variables

const unsigned int GENN_FLAGS::calcSynapseDynamics = 0
 
const unsigned int GENN_FLAGS::calcSynapses = 1
 
const unsigned int GENN_FLAGS::learnSynapsesPost = 2
 
const unsigned int GENN_FLAGS::calcNeurons = 3
 
bool GENN_PREFERENCES::optimiseBlockSize = true
 Flag for signalling whether or not block size optimisation should be performed. More...
 
bool GENN_PREFERENCES::autoChooseDevice = true
 Flag to signal whether the GPU device should be chosen automatically. More...
 
bool GENN_PREFERENCES::optimizeCode = false
 Request speed-optimized code, at the expense of floating-point accuracy. More...
 
bool GENN_PREFERENCES::debugCode = false
 Request debug data to be embedded in the generated code. More...
 
bool GENN_PREFERENCES::showPtxInfo = false
 Request that PTX assembler information be displayed for each CUDA kernel during compilation. More...
 
bool GENN_PREFERENCES::buildSharedLibrary = false
 Should generated code and Makefile build into a shared library e.g. for use in SpineML simulator. More...
 
bool GENN_PREFERENCES::autoInitSparseVars = false
 Previously, variables associated with sparse synapse populations were not automatically initialised. If this flag is set this now occurs in the initMODEL_NAME function and copyStateToDevice is deferred until here. More...
 
bool GENN_PREFERENCES::mergePostsynapticModels = false
 Should compatible postsynaptic models and dendritic delay buffers be merged? This can significantly reduce the cost of updating neuron population but means that per-synapse group inSyn arrays can not be retrieved. More...
 
VarMode GENN_PREFERENCES::defaultVarMode = VarMode::LOC_HOST_DEVICE_INIT_HOST
 What is the default behaviour for model state variables? Historically, everything was allocated on both host AND device and initialised on HOST. More...
 
VarMode GENN_PREFERENCES::defaultSparseConnectivityMode = VarMode::LOC_HOST_DEVICE_INIT_HOST
 
double GENN_PREFERENCES::asGoodAsZero = 1e-19
 What is the default behaviour for sparse synaptic connectivity? Historically, everything was allocated on both the host AND device and initialised on HOST. More...
 
int GENN_PREFERENCES::defaultDevice = 0
 
unsigned int GENN_PREFERENCES::preSynapseResetBlockSize = 32
 default GPU device; used to determine which GPU to use if chooseDevice is 0 (off) More...
 
unsigned int GENN_PREFERENCES::neuronBlockSize = 32
 
unsigned int GENN_PREFERENCES::synapseBlockSize = 32
 
unsigned int GENN_PREFERENCES::learningBlockSize = 32
 
unsigned int GENN_PREFERENCES::synapseDynamicsBlockSize = 32
 
unsigned int GENN_PREFERENCES::initBlockSize = 32
 
unsigned int GENN_PREFERENCES::initSparseBlockSize = 32
 
unsigned int GENN_PREFERENCES::autoRefractory = 1
 Flag for signalling whether spikes are only reported if thresholdCondition changes from false to true (autoRefractory == 1) or spikes are emitted whenever thresholdCondition is true no matter what.%. More...
 
std::string GENN_PREFERENCES::userCxxFlagsWIN = ""
 Allows users to set specific C++ compiler options they may want to use for all host side code (used for windows platforms) More...
 
std::string GENN_PREFERENCES::userCxxFlagsGNU = ""
 Allows users to set specific C++ compiler options they may want to use for all host side code (used for unix based platforms) More...
 
std::string GENN_PREFERENCES::userNvccFlags = ""
 Allows users to set specific nvcc compiler options they may want to use for all GPU code (identical for windows and unix platforms) More...
 
unsigned int neuronBlkSz
 Global variable containing the GPU block size for the neuron kernel. More...
 
unsigned int preSynapseResetBlkSize
 Global variable containing the GPU block size for the reset kernel run before the synapse kernel. More...
 
unsigned int synapseBlkSz
 Global variable containing the GPU block size for the synapse kernel. More...
 
unsigned int learnBlkSz
 Global variable containing the GPU block size for the learn kernel. More...
 
unsigned int synDynBlkSz
 Global variable containing the GPU block size for the synapse dynamics kernel. More...
 
unsigned int initBlkSz
 Global variable containing the GPU block size for the initialization kernel. More...
 
unsigned int initSparseBlkSz
 Global variable containing the GPU block size for the sparse initialization kernel. More...
 
cudaDeviceProp * deviceProp
 
int theDevice
 Global variable containing the currently selected CUDA device's number. More...
 
int deviceCount
 Global variable containing the number of CUDA devices on this host. More...
 
int hostCount
 Global variable containing the number of hosts within the local compute cluster. More...
 

Detailed Description

Global header file containing a few global variables. Part of the code generation section.

Variable Documentation

◆ deviceCount

int deviceCount

Global variable containing the number of CUDA devices on this host.

◆ deviceProp

cudaDeviceProp* deviceProp

◆ hostCount

int hostCount

Global variable containing the number of hosts within the local compute cluster.

◆ initBlkSz

unsigned int initBlkSz

Global variable containing the GPU block size for the initialization kernel.

◆ initSparseBlkSz

unsigned int initSparseBlkSz

Global variable containing the GPU block size for the sparse initialization kernel.

◆ learnBlkSz

unsigned int learnBlkSz

Global variable containing the GPU block size for the learn kernel.

◆ neuronBlkSz

unsigned int neuronBlkSz

Global variable containing the GPU block size for the neuron kernel.

◆ preSynapseResetBlkSize

unsigned int preSynapseResetBlkSize

Global variable containing the GPU block size for the reset kernel run before the synapse kernel.

◆ synapseBlkSz

unsigned int synapseBlkSz

Global variable containing the GPU block size for the synapse kernel.

◆ synDynBlkSz

unsigned int synDynBlkSz

Global variable containing the GPU block size for the synapse dynamics kernel.

◆ theDevice

int theDevice

Global variable containing the currently selected CUDA device's number.