chrischoy / pytorch-custom-cuda-tutorial Goto Github PK

Contributors

Stargazers

Forkers

xinw1012 andrei-pokrovsky shubhampachori12110095 southatsouth xychen9459 baifengbai locussam akhauriyash iovdin felixmonkey sunbau eternalkingdom ai3dvision lxtgh vinayak19th autohe hyzcn hehuiguo dezhili poodarchu juampamuc xdcesc serge-weihao bleachzou3 concept-inversion tokarev-i-v qiqi-helloworld lifebeyondexpectations rlvc vzyknc yuanchunyu nikhil1008 zheng-ningxin roger1993 baowangmath edmontdants bigchou luchangli03 yiweichen04 nirvanalan allentdan ml-edu ruthvik92 guozhibin2014 gakadam rhtm02 wenbaoqiao vuducquan96 kimjiil baronzaomingyan iq-scm dok4everak47 kozaknazar vaibhavbh-0

__global__ void broadcast_sum_kernel(float *a, float *b, int x, int y, int size)
{
    int i = (blockIdx.x + blockIdx.y * gridDim.x) * blockDim.x + threadIdx.x;
    if (i >= size) return;
    int j = i % y;
    i = i / y;
    int k = i % x;
    a[IDX2D(k, j, y)] += b[k];
}