I need to use mkl to implement easy 3D convolution, for instance, my enter information measurement is 10*450*2456, I exploit a 5*5*5 convolution kernel, the output information measurement ought to be 6*446*2452, I exploit these parameters, consult with the official pattern code, but it surely does’t works. I’ve been utilizing the vsldConvExecX command, it  return an unsure integer, I feel the official instance code and documentation for set the step and code directions are very obscure, hope so as to add some content material to clarify extra clearly, thanks! By the way in which,  I might admire it， if I can get the code for my calculation instance.