Top Answers

1. CUDA memory bandwidth when reading a limited number of finite-sized chunks?
   4 votes
2. Is a combination of atomic CAS for 64 and 32 bit ok?
   4 votes
3. CUDA: release lock implemented by atomic operations
   3 votes
4. Armadillo sizeof(arma::Mat) gives different results between GCC and NVCC
   3 votes
5. How to call CUDA api with optional location?
   3 votes