diff --git a/src/gallium/state_trackers/clover/api/transfer.cpp b/src/gallium/state_trackers/clover/api/transfer.cpp
index f704625..5950462 100644
--- a/src/gallium/state_trackers/clover/api/transfer.cpp
+++ b/src/gallium/state_trackers/clover/api/transfer.cpp
@@ -21,6 +21,7 @@
 //
 
 #include <cstring>
+#include <future>
 
 #include "api/util.hpp"
 #include "core/event.hpp"
@@ -238,24 +239,28 @@ namespace {
                 T dst_obj, const vector_t &dst_orig, const vector_t &dst_pitch,
                 S src_obj, const vector_t &src_orig, const vector_t &src_pitch,
                 const vector_t &region) {
-      return [=, &q](event &) {
-         auto dst = _map<T>::get(q, dst_obj, CL_MAP_WRITE,
-                                 dot(dst_pitch, dst_orig),
-                                 size(dst_pitch, region));
-         auto src = _map<S>::get(q, src_obj, CL_MAP_READ,
-                                 dot(src_pitch, src_orig),
-                                 size(src_pitch, region));
-         vector_t v = {};
-
-         for (v[2] = 0; v[2] < region[2]; ++v[2]) {
-            for (v[1] = 0; v[1] < region[1]; ++v[1]) {
-               std::memcpy(
-                  static_cast<char *>(dst) + dot(dst_pitch, v),
-                  static_cast<const char *>(src) + dot(src_pitch, v),
-                  src_pitch[0] * region[0]);
-            }
-         }
-      };
+      auto done = static_cast<std::shared_future<void>>(
+         std::async(std::launch::async, [=, &q]() {
+                auto dst = _map<T>::get(q, dst_obj, CL_MAP_WRITE,
+                                        dot(dst_pitch, dst_orig),
+                                        size(dst_pitch, region));
+                auto src = _map<S>::get(q, src_obj, CL_MAP_READ,
+                                        dot(src_pitch, src_orig),
+                                        size(src_pitch, region));
+                vector_t v = {};
+
+                for (v[2] = 0; v[2] < region[2]; ++v[2]) {
+                   for (v[1] = 0; v[1] < region[1]; ++v[1]) {
+                      std::memcpy(
+                         static_cast<char *>(dst) + dot(dst_pitch, v),
+                         static_cast<const char *>(src) + dot(src_pitch, v),
+                         src_pitch[0] * region[0]);
+                   }
+                }
+             }
+          )
+      );
+      return [done](event &) { done.wait(); };
    }
 
    ///