1
vote
1answer
304 views

Cuda driver, CPU/GPU performances issue [closed]

I implemented a RNS Montgomery exponentiation in CUDA and on CPU for comparison. Everything nice everything fine. It runs on just one SM. However, I am going to tell you some strange regression in ...
1
vote
0answers
50 views

Why is rvalue write in shared memory array serialised? [closed]

I'm using CUDA 4.0 on a GPU with computing capability 2.1. One of my device functions is the following: device void test(int n, int* itemp) // itemp is shared memory pointer { const int tid = ...