Multithreading in TVM

Hello! Is it possible to use TVM objects in multiple threads with CUDA?