How to fix the CUDA kernel error of 0 valued outputs

manpreet Best Answer 2 years ago

This is the code for a simple b.com/tag/n">neural b.com/tag/n">network coded ib.com/tag/n">n c++ usib.com/tag/n">ng visual studio, I have eb.com/tag/n">ncoub.com/tag/n">ntered a rather commob.com/tag/n">n CUDA b.com/tag/error">error for this code

#ib.com/tag/n">nclude "cuda_rub.com/tag/n">ntime.h"
#ib.com/tag/n">nclude "device_laub.com/tag/n">nch_parameters.h"
#ib.com/tag/n">nclude 
#ib.com/tag/n">nclude 
#ib.com/tag/n">nclude 
#ib.com/tag/n">nclude b.com/tag/n">ng>
#ib.com/tag/n">nclude 
#ib.com/tag/n">nclude b.com/tag/n">ng.h>
#ib.com/tag/n">nclude 
#defib.com/tag/n">ne LRATE 0.0001
#defib.com/tag/n">ne BIAS 0.0
#defib.com/tag/n">ne DEFAULTFILENAME "Default.dat"
#defib.com/tag/n">ne LAYER1SIZE 2000
#defib.com/tag/n">ne LAYER2SIZE 1000
#defib.com/tag/n">ne LAYER3SIZE 200
#defib.com/tag/n">ne LAYER4SIZE 100
#defib.com/tag/n">ne LAYER5SIZE 20
#defib.com/tag/n">ne LAYER6SIZE 10
#defib.com/tag/n">ne LAYER7SIZE 3
char FName[b.com/tag/30">30];
````Cuda kerb.com/tag/n">nels````
__global__ void mulMat(float *A, float *B, float *C, ib.com/tag/n">nt N, ib.com/tag/n">nt Nsize)
{
    C[threadIdx.x] = A[threadIdx.x*Nsize + N] * B[N];
}
__global__ void errPush(float *A, float *B, float *C, ib.com/tag/n">nt N, ib.com/tag/n">nt Nsize)
{
    C[threadIdx.x] = A[N*blockDim.x+threadIdx.x]*B[N];

}
__global__ void gradPush(float *ErrorDef, float *ActivatedVal, float *GradMat)
{
    GradMat[threadIdx.x*blockIdx.y] = ErrorDef[threadIdx.x] * ActivatedVal[threadIdx.y];

}
__global__ void gradieb.com/tag/n">ntSub(float *A, float *B)
{
    A[threadIdx.x*blockDim.y + threadIdx.y] = A[threadIdx.x*blockDim.y + threadIdx.y] <