slightly modify the debug benchmark descriptor to contain multiple threads and also a batch

benchmarks/benchmark-descriptors/copy-debug-n0ton0-cpu.json

@@ -1,5 +1,5 @@
 {
-  "count": 1,
+  "count": 2,
   "list": [
     {
       "affinity": {
@@ -11,6 +11,17 @@
         "size": 1024,
         "batch_size": 0
       }
+    },
+    {
+      "affinity": {
+        "nnode_dst": 0,
+        "nnode_src": 0,
+        "node": 0
+      },
+      "task": {
+        "size": 1024,
+        "batch_size": 10
+      }
     }
   ],
   "path": "sw",