diff --git a/ChASE-MPI/mpi_wrapper.hpp b/ChASE-MPI/mpi_wrapper.hpp
index fc77569..0cfca27 100644
--- a/ChASE-MPI/mpi_wrapper.hpp
+++ b/ChASE-MPI/mpi_wrapper.hpp
@@ -150,6 +150,7 @@ void Memcpy(int mode, void* dst, const void* src, std::size_t count)
 #if defined(CUDA_AWARE)
         case CPY_D2D:
             cudaMemcpy(dst, src, count, cudaMemcpyDeviceToDevice);
+            cudaDeviceSynchronize();
             break;
         case CPY_D2H:
             cudaMemcpy(dst, src, count, cudaMemcpyDeviceToHost);