#include <stdio.h> #include <iostream> #include <math.h> // Kernel function to add the elements of two arrays #define I 7 #define gna 35. #define gk 9. #define gl 0.1 #define ena 55. #define ek (-90.) #define el (-65.) #define dt 0.01 // Set GPU parallelization #define BLOCKS 4 #define THREADS 256 // Set simulation time #define TIME_ITERATIONS 6000000l __global__ void run(float *v, float *h, float *n) { int i = blockIdx.x * blockDim.x + threadIdx.x; float minf, ninf, hinf, ntau, htau; for(unsigned long t = 0; t<TIME_ITERATIONS; ++t){ minf = (0.0313077505826+0.000327941702344*v[i])+\ ((v[i]>-63.8824871877)?(0.984216321118+0.0154066687827*v[i]):0.)+\ ((v[i]>-6.79083019467)?(-0.0960057625861+-0.0141375590074*v[i]):0.); ninf = (0.0964837408311+0.00095221890206*v[i])+\ ((v[i]>-71.0303053169)?(0.743014681244+0.0104605305852*v[i]):0.)+\ ((v[i]>-4.02151450044)?(-0.0343121915059+-0.00853215660471*v[i]):0.); ntau = (3.89986986823+0.0365576715394*v[i])+\ ((v[i]>-60.3962620374)?(-4.73857532036+-0.0784580893008*v[i]):0.)+\ ((v[i]>-26.7339539639)?(1.09061524502+0.0407951344007*v[i]):0.); hinf = (0.961765215359+-0.000371676284993*v[i])+\ ((v[i]>-81.574862488)?(-1.32606283442+-0.0162557777479*v[i]):0.)+\ ((v[i]>-25.0499306576)?(0.397064011884+0.0158509026357*v[i]):0.); htau = (1.78046831377+0.00985969575028*v[i])+\ ((v[i]>-57.5852330206)?(-1.29845989067+-0.0225484872173*v[i]):0.)+\ ((v[i]>5.61104313787)?(-0.0499171171021+0.00889622764887*v[i]):0.); n[i] += dt*(ninf - n[i])/ntau; h[i] += dt*(hinf - h[i])/htau; v[i] += dt*(-gna*minf*minf*minf*h[i]*(v[i]-ena)-gk*n[i]*n[i]*n[i]*n[i]*(v[i]-ek)-gl*(v[i]-el)+I); //printf("%g %g\n",dt*t,v); } } int main(void) { int N = BLOCKS*THREADS; float *v, *h, *n; // Allocate Unified Memory – accessible from CPU or GPU cudaMallocManaged(&v, N*sizeof(float)); cudaMallocManaged(&h, N*sizeof(float)); cudaMallocManaged(&n, N*sizeof(float)); // initialize arrays on the host for (int i = 0; i < N; i++) { v[i] = -63.f; h[i] = n[i] = 0.f; } // Run kernel on the GPU run<<<BLOCKS, THREADS>>>(v, h, n); // Wait for GPU to finish before accessing on host cudaDeviceSynchronize(); //check for errors cudaError_t e = cudaGetLastError(); if(e){ printf("ERROR (%d): %s\n",e,cudaGetErrorString(e)); } // Free memory cudaFree(v); cudaFree(h); cudaFree(n); return 0; }