I have posted the same post in Nvidia CUDA forum: Link Here I reinstalled multiple times with the method from the post but still having the same “class template” problems. Then I reinstalled CUDA 9.1 and VS2017 ver 15.6.7 also with the same method and it finally works. Further problems that I encountered is in … Read more
I think you’re really trying to ask about branchless vs. branching ways to do m = max(m, array[i]). C compilers will already compile the if() version to branchless code (using cmov) depending on optimization settings. It can even auto-vectorize to a packed compare or packed-max function. Your 0.5 * abs() version is obviously terrible (a … Read more
Your example code should not give any errors as was proved by Robert Crovella. I looked at your original code and found a misaligned parenthesis cause by that first “for” loop. Good luck… 0 solved CUDA: Using a global thread index with shared memory won’t work [closed]
If you have experience with Matlab, the easiest path is to go through the “VGG Convolutional Neural Networks Practical” and use their open source MatConvNet toolbox for Matlab: http://www.vlfeat.org/matconvnet/. 0 solved Neural Network Image Classification, The Most Efficient Solution / Suggestion [closed]