diff --git a/examples_cuda/sort/sort b/examples_cuda/sort/sort
index d148e465..64c32dd0 100755
Binary files a/examples_cuda/sort/sort and b/examples_cuda/sort/sort differ
diff --git a/examples_cuda/sort/sort.cpp b/examples_cuda/sort/sort.cpp
index 37a5c289..8e3081e5 100644
--- a/examples_cuda/sort/sort.cpp
+++ b/examples_cuda/sort/sort.cpp
@@ -40,6 +40,7 @@
 #include <iomanip>
 #include "../timing.h"
 #include "sort_ispc.h"
+#include <cassert>
 
 #include <sys/time.h>
 static inline double rtc(void)
@@ -129,6 +130,12 @@ int main (int argc, char *argv[])
 
   printf("[sort ispc + tasks]:\t[%.3f] million cycles\n", tISPC2);
 
+  for (int i = 0; i < n-1; i++)
+  {
+    assert(code[i+1] >=  code[i]);
+  }
+
+
   srand (0);
 
   for (i = 0; i < m; i ++)
diff --git a/examples_cuda/sort/sort_cu.cpp b/examples_cuda/sort/sort_cu.cpp
index ec295434..b8e9edf5 100644
--- a/examples_cuda/sort/sort_cu.cpp
+++ b/examples_cuda/sort/sort_cu.cpp
@@ -351,7 +351,7 @@ int main (int argc, char *argv[])
   createContext();
   /*******************/
 
-  int ntask = 13*4*8;
+  int ntask = 13*4;
   devicePtr d_code   = deviceMalloc(n*sizeof(int));
   devicePtr d_order  = deviceMalloc(n*sizeof(int));
   devicePtr d_pair   = deviceMalloc(n*2*sizeof(int));
@@ -386,7 +386,13 @@ int main (int argc, char *argv[])
         progressbar (i, m);
   }
 
-  printf("[sort ispc + tasks]:\t[%.3f] million cycles\n", tISPC2);
+  printf("[sort cuda]:\t[%.3f] million cycles\n", tISPC2);
+  memcpyD2H(code,  d_code,  n*sizeof(int));
+  memcpyD2H(order, d_order, n*sizeof(int));
+  for (int i = 0; i < n-1; i++)
+  {
+    assert(code[i+1] >=  code[i]);
+  }
 
   srand (0);