diff --git a/02-MSA-hello-world-gpu/.master/hello-world.cu b/02-MSA-hello-world-gpu/.master/hello-world.cu
index ad3ee874b045a2a9d313169ce78bf032e3a6cb1e..1194b5cc20c0f8966465b0593e60b283bc043df4 100644
--- a/02-MSA-hello-world-gpu/.master/hello-world.cu
+++ b/02-MSA-hello-world-gpu/.master/hello-world.cu
@@ -51,7 +51,7 @@ int main(int argc, char** argv){
     CUDA_RT_CALL( cudaPeekAtLastError() );
     CUDA_RT_CALL( cudaDeviceSynchronize() );
     
-    // TODO: Implement the MPI_Recv() call to receive the "payload" from rank 1 using directly "d_payload" as the target buffer on the GPU
+    // TODO: Implement the MPI_Recv() call to receive the "payload" from rank 0 using directly "d_payload" as the target buffer on the GPU
     if (rank == 1) {
     #ifdef SOLUTION
         MPI_Recv(d_payload, 6, MPI_CHAR, 0, tag, MPI_COMM_WORLD, &stat);
diff --git a/02-MSA-hello-world-gpu/solutions/hello-world.cu b/02-MSA-hello-world-gpu/solutions/hello-world.cu
index 77ebbc60bfb71df0a2ed4427816d55a0e7c66259..5f22101926e05a3c4f8081f92fabe390f3710201 100644
--- a/02-MSA-hello-world-gpu/solutions/hello-world.cu
+++ b/02-MSA-hello-world-gpu/solutions/hello-world.cu
@@ -51,7 +51,7 @@ int main(int argc, char** argv){
     CUDA_RT_CALL( cudaPeekAtLastError() );
     CUDA_RT_CALL( cudaDeviceSynchronize() );
     
-    // TODO: Implement the MPI_Recv() call to receive the "payload" from rank 1 using directly "d_payload" as the target buffer on the GPU
+    // TODO: Implement the MPI_Recv() call to receive the "payload" from rank 0 using directly "d_payload" as the target buffer on the GPU
     if (rank == 1) {
         MPI_Recv(d_payload, 6, MPI_CHAR, 0, tag, MPI_COMM_WORLD, &stat);
     }
diff --git a/02-MSA-hello-world-gpu/tasks/hello-world.cu b/02-MSA-hello-world-gpu/tasks/hello-world.cu
index 17103d4bdedffea023aae35ca0b36fcd5af2515e..b763dacf3150e442608c851f05ba1b07da901bdf 100644
--- a/02-MSA-hello-world-gpu/tasks/hello-world.cu
+++ b/02-MSA-hello-world-gpu/tasks/hello-world.cu
@@ -51,7 +51,7 @@ int main(int argc, char** argv){
     CUDA_RT_CALL( cudaPeekAtLastError() );
     CUDA_RT_CALL( cudaDeviceSynchronize() );
     
-    // TODO: Implement the MPI_Recv() call to receive the "payload" from rank 1 using directly "d_payload" as the target buffer on the GPU
+    // TODO: Implement the MPI_Recv() call to receive the "payload" from rank 0 using directly "d_payload" as the target buffer on the GPU
     if (rank == 1) {
         MPI_Recv();
     }