diff --git a/cscs-checks/mch/src/Makefile b/cscs-checks/mch/src/Makefile
index 5de7f6f0be..e8bd695e03 100644
--- a/cscs-checks/mch/src/Makefile
+++ b/cscs-checks/mch/src/Makefile
@@ -6,7 +6,7 @@ LD = $(FC)
 
 OBJS = compute_cuda.o openacc_cuda_mpi_cppstd.o std_cpp_call.o
 # OBJ2 = $(subst _,$(PE_ENV)_,$(OBJ))
-LIB  =
+LIB  =  -lcuda -lcudart -lstdc++
 
 .SUFFIXES: .o .cu .cpp .F90
 
diff --git a/cscs-checks/mch/src/compute_cuda.cu b/cscs-checks/mch/src/compute_cuda.cu
index bd77a2d23c..74a58a7fe9 100644
--- a/cscs-checks/mch/src/compute_cuda.cu
+++ b/cscs-checks/mch/src/compute_cuda.cu
@@ -1,5 +1,6 @@
 #include  <stdio.h>
 #include "cuda.h"
+#include "mpi.h"
 
 #define cudaCheckErrors(msg) \
     do { \
@@ -35,10 +36,34 @@ void cuda_kernel_no_copy(float* a, float* b, int n)
   cudaCheckErrors("cuda error");
 }
 
+void mpi_hello_world(int comm)
+{
+  MPI_Init(NULL, NULL);
+  MPI_Comm comm2; 
+  MPI_Comm_dup(MPI_Comm_f2c(comm), &comm2);
+
+  // Get the rank of the process
+  int world_size;
+  MPI_Comm_size(comm2, &world_size);
+
+  int world_rank;
+  MPI_Comm_rank(comm2, &world_rank);
+
+  // Get the name of the processor
+  char processor_name[MPI_MAX_PROCESSOR_NAME];
+  int name_len;
+  MPI_Get_processor_name(processor_name, &name_len);
+
+  // Print off a hello world message
+  printf("Hallo world from processor %s, rank %d out of %d processors\n",
+         processor_name, world_rank, world_size);
+
+}
 void cuda_kernel_with_copy(float* a, float* b, int n)
 {
   const int THREADS_PER_BLOCK = 1;
   const int NUMBER_OF_BLOCKS = 10;
+ 
 
   float* d_a;
   float* d_b;
diff --git a/cscs-checks/mch/src/openacc_cuda_mpi_cppstd.F90 b/cscs-checks/mch/src/openacc_cuda_mpi_cppstd.F90
index 4568b9f4f5..2e81680440 100644
--- a/cscs-checks/mch/src/openacc_cuda_mpi_cppstd.F90
+++ b/cscs-checks/mch/src/openacc_cuda_mpi_cppstd.F90
@@ -35,7 +35,6 @@ program openacc_cuda_mpi_cppstd
   !$acc end host_data
   !$acc end data
 
-
   if(mpi_rank == 0) then
 
     ! Allocate and initialize arrays on the GPU
@@ -75,9 +74,9 @@ program openacc_cuda_mpi_cppstd
     if (sum(f1) /= EXPECTED_CUDA_SUM) then
       write (*,*) "Result : FAIL"
       write (*,*) "Expected value sum(f1): ", EXPECTED_CUDA_SUM, "actual value:", sum(f1)
-    else if (sum(f3) /= EXPECTED_CUDA_SUM) then
+    else if (sum(f3) /= EXPECTED_CPP_STD_SUM) then
       write (*,*) "Result : FAIL"
-      write (*,*) "Expected value sum(f3): ", EXPECTED_CUDA_SUM, "actual value:", sum(f3)
+      write (*,*) "Expected value sum(f3): ", EXPECTED_CPP_STD_SUM, "actual value:", sum(f3)
     else if (data_sum(1) /= ref_val) then
       write (*,*) "Result : FAIL"
       write (*,*) "Expected value data_sum: ", ref_val, "actual value:", data_sum(1)
@@ -93,7 +92,6 @@ program openacc_cuda_mpi_cppstd
     deallocate(f1)
     deallocate(f2)
     deallocate(f3)
-    write (*,*) "Result: OK"
   end if
 
   call MPI_Finalize(ierr);