efield.cuh 11.5 KB
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409
#ifndef	RTS_EFIELD
#define RTS_EFIELD

#include "../math/complex.h"
#include "../math/realfield.cuh"
#include "../visualization/colormap.h"
#include "../optics/planewave.h"
#include "../cuda/devices.h"
#include "../optics/beam.h"
#include "../math/rect.h"


namespace stim{

template<typename T>
__global__ void gpu_planewave2efield(complex<T>* X, complex<T>* Y, complex<T>* Z, unsigned int r0, unsigned int r1, 
									 planewave<T> w, rect<T> q)
{
    int iu = blockIdx.x * blockDim.x + threadIdx.x;
    int iv = blockIdx.y * blockDim.y + threadIdx.y;

    //make sure that the thread indices are in-bounds
    if(iu >= r0 || iv >= r1) return;

    //compute the index into the field
    int i = iv*r0 + iu;

	//get the current position
	vec<T> p = q( (T)iu/(T)r0, (T)iv/(T)r1 );
	vec<T> r(p[0], p[1], p[2]);

	complex<T> x( 0.0f, w.k.dot(r) );

    if(Y == NULL)                       //if this is a scalar simulation
        X[i] += w.E0.len() * exp(x);    //use the vector magnitude as the plane wave amplitude
    else
    {
        X[i] += w.E0[0] * exp(x);
        Y[i] += w.E0[1] * exp(x);
        Z[i] += w.E0[2] * exp(x);
    }
}

template<typename T>
__global__ void gpu_efield_magnitude(complex<T>* X, complex<T>* Y, complex<T>* Z, unsigned int r0, unsigned int r1, T* M)
{
	int iu = blockIdx.x * blockDim.x + threadIdx.x;
    int iv = blockIdx.y * blockDim.y + threadIdx.y;

    //make sure that the thread indices are in-bounds
    if(iu >= r0 || iv >= r1) return;

    //compute the index into the field
    int i = iv*r0 + iu;

    if(Y == NULL)                      //if this is a scalar simulation
	   M[i] = X[i].abs();              //compute the magnitude of the X component
    else
    {
        M[i] = rts::vec<T>(X[i].abs(), Y[i].abs(), Z[i].abs()).len();
        //M[i] = Z[i].abs();
    }
}

template<typename T>
__global__ void gpu_efield_real_magnitude(complex<T>* X, complex<T>* Y, complex<T>* Z, unsigned int r0, unsigned int r1, T* M)
{
    int iu = blockIdx.x * blockDim.x + threadIdx.x;
    int iv = blockIdx.y * blockDim.y + threadIdx.y;

    //make sure that the thread indices are in-bounds
    if(iu >= r0 || iv >= r1) return;

    //compute the index into the field
    int i = iv*r0 + iu;

    if(Y == NULL)                      //if this is a scalar simulation
       M[i] = X[i].abs();              //compute the magnitude of the X component
    else
    {
        M[i] = rts::vec<T>(X[i].real(), Y[i].real(), Z[i].real()).len();
        //M[i] = Z[i].abs();
    }
}

template<typename T>
__global__ void gpu_efield_polarization(complex<T>* X, complex<T>* Y, complex<T>* Z, 
                                        unsigned int r0, unsigned int r1,
                                        T* Px, T* Py, T* Pz)
{
    int iu = blockIdx.x * blockDim.x + threadIdx.x;
    int iv = blockIdx.y * blockDim.y + threadIdx.y;

    //make sure that the thread indices are in-bounds
    if(iu >= r0 || iv >= r1) return;

    //compute the index into the field
    int i = iv*r0 + iu;

    //compute the field polarization
    Px[i] = X[i].abs();
    Py[i] = Y[i].abs();
    Pz[i] = Z[i].abs();

}

/*	This function computes the sum of two complex fields and stores the result in *dest
*/
template<typename T>
__global__ void gpu_efield_sum(complex<T>* dest, complex<T>* src, unsigned int r0, unsigned int r1)
{
	int iu = blockIdx.x * blockDim.x + threadIdx.x;
    int iv = blockIdx.y * blockDim.y + threadIdx.y;

    //make sure that the thread indices are in-bounds
    if(iu >= r0 || iv >= r1) return;

    //compute the index into the field
    int i = iv*r0 + iu;

    //sum the fields
    dest[i] += src[i];
}

/*  This class implements a discrete representation of an electromagnetic field
    in 2D. The majority of this representation is done on the GPU.
*/
template<typename T>
class efield
{
protected:

    bool vector;

    //gpu pointer to the field data
    rts::complex<T>* X;
    rts::complex<T>* Y;
    rts::complex<T>* Z;

    //resolution of the discrete field
    unsigned int R[2];

    //shape of the 2D field
    rect<T> pos;

	void from_planewave(planewave<T> p)
	{
        unsigned int SQRT_BLOCK = 16;
        //create one thread for each detector pixel
        dim3 dimBlock(SQRT_BLOCK, SQRT_BLOCK);
        dim3 dimGrid((R[0] + SQRT_BLOCK -1)/SQRT_BLOCK, (R[1] + SQRT_BLOCK - 1)/SQRT_BLOCK);
        

        gpu_planewave2efield<T> <<<dimGrid, dimBlock>>> (X, Y, Z, R[0], R[1], p, pos);
    }

    void init(unsigned int res0, unsigned int res1, bool _vector)
    {
    	vector = _vector;           //initialize field type
        
        X = Y = Z = NULL;           //initialize all pointers to NULL
        R[0] = res0;
        R[1] = res1;

        //allocate memory on the gpu
        cudaMalloc(&X, sizeof(rts::complex<T>) * R[0] * R[1]);
		cudaMemset(X, 0, sizeof(rts::complex<T>) * R[0] * R[1]);

        if(vector)
        {
            cudaMalloc(&Y, sizeof(rts::complex<T>) * R[0] * R[1]);
			cudaMemset(Y, 0, sizeof(rts::complex<T>) * R[0] * R[1]);

            cudaMalloc(&Z, sizeof(rts::complex<T>) * R[0] * R[1]);
			cudaMemset(Z, 0, sizeof(rts::complex<T>) * R[0] * R[1]);
        }
    }

    void destroy()
    {
		if(X != NULL) cudaFree(X);
		if(Y != NULL) cudaFree(Y);
		if(Z != NULL) cudaFree(Z);
    }

    void shallowcpy(const rts::efield<T> & src)
    {
    	vector = src.vector;
    	R[0] = src.R[0];
    	R[1] = src.R[1];
    }

    void deepcpy(const rts::efield<T> & src)
    {
    	//perform a shallow copy
    	shallowcpy(src);

    	//allocate memory on the gpu
    	if(src.X != NULL)
    	{
    		cudaMalloc(&X, sizeof(rts::complex<T>) * R[0] * R[1]);
    		cudaMemcpy(X, src.X, sizeof(rts::complex<T>) * R[0] * R[1], cudaMemcpyDeviceToDevice);
    	}
    	if(src.Y != NULL)
    	{
    		cudaMalloc(&Y, sizeof(rts::complex<T>) * R[0] * R[1]);
    		cudaMemcpy(Y, src.Y, sizeof(rts::complex<T>) * R[0] * R[1], cudaMemcpyDeviceToDevice);
    	}
    	if(src.Z != NULL)
    	{
    		cudaMalloc(&Z, sizeof(rts::complex<T>) * R[0] * R[1]);
    		cudaMemcpy(Z, src.Z, sizeof(rts::complex<T>) * R[0] * R[1], cudaMemcpyDeviceToDevice);
    	}
    }

public:
    efield(unsigned int res0, unsigned int res1, bool _vector = true)
    {
        init(res0, res1, _vector);
        //pos = rts::rect<T>(rts::vec<T>(-10, 0, -10), rts::vec<T>(-10, 0, 10), rts::vec<T>(10, 0, 10));
    }

    //destructor
    ~efield()
    {
    	destroy();
    }

    ///Clear the field - set all points to zero
    void clear()
    {
        cudaMemset(X, 0, sizeof(rts::complex<T>) * R[0] * R[1]);

        if(vector)
        {
            cudaMemset(Y, 0, sizeof(rts::complex<T>) * R[0] * R[1]);
            cudaMemset(Z, 0, sizeof(rts::complex<T>) * R[0] * R[1]);
        }
    }

    void position(rect<T> _p)
    {
        pos = _p;
    }

    //access functions
    complex<T>* x(){
        return X;
    }
    complex<T>* y(){
        return Y;
    }
    complex<T>* z(){
        return Z;
    }
    unsigned int Ru(){
        return R[0];
    }
    unsigned int Rv(){
        return R[1];
    }
    rect<T> p(){
        return pos;
    }

    std::string str()
    {
        stringstream ss;
        ss<<pos<<std::endl;
        ss<<"X: "<<X<<std::endl;
        ss<<"Y: "<<Y<<std::endl;
        ss<<"Z: "<<Z;

        return ss.str();
    }

    //assignment operator: assignment from another electric field
    efield<T> & operator= (const efield<T> & rhs)
    {
    	destroy();				//destroy any previous information about this field
    	deepcpy(rhs);			//create a deep copy
    	return *this;			//return the current object
    }

    //assignment operator: build an electric field from a plane wave
    efield<T> & operator= (const planewave<T> & rhs)
	{
		
		clear();				//clear any previous field data
		from_planewave(rhs);	//create a field from the planewave
		return *this;			//return the current object
	}

	//assignment operator: add an existing electric field
	efield<T> & operator+= (const efield<T> & rhs)
	{
		//if this field and the source field represent the same regions in space
		if(R[0] == rhs.R[0] && R[1] == rhs.R[1] && pos == rhs.pos)
		{
			int maxThreads = rts::maxThreadsPerBlock(); //compute the optimal block size
			int SQRT_BLOCK = (int)std::sqrt((float)maxThreads);

			//create one thread for each detector pixel
			dim3 dimBlock(SQRT_BLOCK, SQRT_BLOCK);
			dim3 dimGrid((R[0] + SQRT_BLOCK -1)/SQRT_BLOCK, (R[1] + SQRT_BLOCK - 1)/SQRT_BLOCK);

			//sum the fields
			gpu_efield_sum <<<dimGrid, dimBlock>>> (X, rhs.X, R[0], R[1]);
			if(Y != NULL)
			{
				gpu_efield_sum <<<dimGrid, dimBlock>>> (Y, rhs.Y, R[0], R[1]);
				gpu_efield_sum <<<dimGrid, dimBlock>>> (Z, rhs.Z, R[0], R[1]);
			}
		}
		else
		{
			std::cout<<"ERROR in efield: The two summed fields do not represent the same geometry."<<std::endl;
			exit(1);
		}

		return *this;		//return this object
	}

    //assignment operator: build an electric field from a beam
    efield<T> & operator= (const rts::beam<T> & rhs)
    {
        //get a vector of monte-carlo samples
        std::vector< rts::planewave<T> > p_list = rhs.mc();

        clear();                //clear any previous field data
        for(unsigned int i = 0; i < p_list.size(); i++)
            from_planewave(p_list[i]);
        return *this;
    }


	//return a scalar field representing field magnitude
    realfield<T, 1, true> mag()
    {
		int maxThreads = rts::maxThreadsPerBlock(); //compute the optimal block size
		int SQRT_BLOCK = (int)std::sqrt((float)maxThreads);

		//create a scalar field to store the result
		realfield<T, 1, true> M(R[0], R[1]);

		//create one thread for each detector pixel
		dim3 dimBlock(SQRT_BLOCK, SQRT_BLOCK);
		dim3 dimGrid((R[0] + SQRT_BLOCK -1)/SQRT_BLOCK, (R[1] + SQRT_BLOCK - 1)/SQRT_BLOCK);

		//compute the magnitude and store it in a scalar field
		gpu_efield_magnitude<float> <<<dimGrid, dimBlock>>> (X, Y, Z, R[0], R[1], M.ptr(0));

		return M;
    }

    //return a sacalar field representing the field magnitude at an infinitely small point in time
    realfield<T, 1, true> real_mag()
    {
        int maxThreads = rts::maxThreadsPerBlock(); //compute the optimal block size
        int SQRT_BLOCK = (int)std::sqrt((float)maxThreads);

        //create a scalar field to store the result
        realfield<T, 1, true> M(R[0], R[1]);

        //create one thread for each detector pixel
        dim3 dimBlock(SQRT_BLOCK, SQRT_BLOCK);
        dim3 dimGrid((R[0] + SQRT_BLOCK -1)/SQRT_BLOCK, (R[1] + SQRT_BLOCK - 1)/SQRT_BLOCK);

        //compute the magnitude and store it in a scalar field
        gpu_efield_real_magnitude<float> <<<dimGrid, dimBlock>>> (X, Y, Z, R[0], R[1], M.ptr(0));

        return M;
    }

    //return a vector field representing field polarization
    realfield<T, 3, true> polarization()
    {
        if(!vector)
        {
            std::cout<<"ERROR: Cannot compute polarization of a scalar field."<<std::endl;
            exit(1);
        }
        int maxThreads = rts::maxThreadsPerBlock(); //compute the optimal block size
        int SQRT_BLOCK = (int)std::sqrt((float)maxThreads);        

        //create one thread for each detector pixel
        dim3 dimBlock(SQRT_BLOCK, SQRT_BLOCK);
        dim3 dimGrid((R[0] + SQRT_BLOCK -1)/SQRT_BLOCK, (R[1] + SQRT_BLOCK - 1)/SQRT_BLOCK);

        
        realfield<T, 3, true> Pol(R[0], R[1]);     //create a vector field to store the result

        //compute the polarization and store it in the vector field
        gpu_efield_polarization<float> <<<dimGrid, dimBlock>>> (X, Y, Z, R[0], R[1], Pol.ptr(0), Pol.ptr(1), Pol.ptr(2));

        return Pol;                         //return the vector field
    }

    ////////FRIENDSHIP
    //friend void operator<< <T>(rts::efield<T> &ef, rts::halfspace<T> hs);


};



}   //end namespace rts

#endif