diff --git a/openmp/libomptarget/include/omptarget.h b/openmp/libomptarget/include/omptarget.h
--- a/openmp/libomptarget/include/omptarget.h
+++ b/openmp/libomptarget/include/omptarget.h
@@ -168,7 +168,7 @@
   DeviceTy &Device;
 
 public:
-  AsyncInfoTy(DeviceTy &Device) : Device(Device) {}
+  AsyncInfoTy(DeviceTy &Device);
   ~AsyncInfoTy() { synchronize(); }
 
   /// Implicit conversion to the __tgt_async_info which is used in the
diff --git a/openmp/libomptarget/include/omptargetplugin.h b/openmp/libomptarget/include/omptargetplugin.h
--- a/openmp/libomptarget/include/omptargetplugin.h
+++ b/openmp/libomptarget/include/omptargetplugin.h
@@ -70,6 +70,11 @@
 void *__tgt_rtl_data_alloc(int32_t ID, int64_t Size, void *HostPtr,
                            int32_t Kind);
 
+// Create object for device synchronization. In case of success, return zero
+// and AsyncInfo->Queue is not a nullptr. On failure, return an error code and
+// set AsyncInfo->Queue to nullptr.
+int32_t __tgt_rtl_create_async_info(int32_t ID, __tgt_async_info *AsyncInfo);
+
 // Pass the data content to the target device using the target address. In case
 // of success, return zero. Otherwise, return an error code.
 int32_t __tgt_rtl_data_submit(int32_t ID, void *TargetPtr, void *HostPtr,
diff --git a/openmp/libomptarget/plugins/amdgpu/src/rtl.cpp b/openmp/libomptarget/plugins/amdgpu/src/rtl.cpp
--- a/openmp/libomptarget/plugins/amdgpu/src/rtl.cpp
+++ b/openmp/libomptarget/plugins/amdgpu/src/rtl.cpp
@@ -2204,6 +2204,11 @@
   return ptr;
 }
 
+int32_t __tgt_rtl_create_data_async(int, __tgt_async_info *AsyncInfo) {
+  initAsyncInfo(AsyncInfo);
+  return OFFLOAD_SUCCESS;
+}
+
 int32_t __tgt_rtl_data_submit(int device_id, void *tgt_ptr, void *hst_ptr,
                               int64_t size) {
   assert(device_id < DeviceInfo.NumberOfDevices && "Device ID too large");
diff --git a/openmp/libomptarget/plugins/cuda/src/rtl.cpp b/openmp/libomptarget/plugins/cuda/src/rtl.cpp
--- a/openmp/libomptarget/plugins/cuda/src/rtl.cpp
+++ b/openmp/libomptarget/plugins/cuda/src/rtl.cpp
@@ -468,15 +468,6 @@
     E.Table.EntriesBegin = E.Table.EntriesEnd = nullptr;
   }
 
-  CUstream getStream(const int DeviceId, __tgt_async_info *AsyncInfo) const {
-    assert(AsyncInfo && "AsyncInfo is nullptr");
-
-    if (!AsyncInfo->Queue)
-      AsyncInfo->Queue = StreamManager->getStream(DeviceId);
-
-    return reinterpret_cast<CUstream>(AsyncInfo->Queue);
-  }
-
 public:
   // This class should not be copied
   DeviceRTLTy(const DeviceRTLTy &) = delete;
@@ -950,6 +941,15 @@
     return nullptr;
   }
 
+  CUstream getStream(const int DeviceId, __tgt_async_info *AsyncInfo) const {
+    assert(AsyncInfo && "AsyncInfo is nullptr");
+
+    if (!AsyncInfo->Queue)
+      AsyncInfo->Queue = StreamManager->getStream(DeviceId);
+
+    return reinterpret_cast<CUstream>(AsyncInfo->Queue);
+  }
+
   int dataSubmit(const int DeviceId, const void *TgtPtr, const void *HstPtr,
                  const int64_t Size, __tgt_async_info *AsyncInfo) const {
     assert(AsyncInfo && "AsyncInfo is nullptr");
@@ -1450,6 +1450,11 @@
   return DeviceRTL.dataAlloc(device_id, size, (TargetAllocTy)kind);
 }
 
+int32_t __tgt_rtl_create_data_async(int DeviceId, __tgt_async_info *AsyncInfo) {
+  DeviceRTL.getStream(DeviceId, AsyncInfo);
+  return AsyncInfo->Queue ? OFFLOAD_SUCCESS : OFFLOAD_FAIL;
+}
+
 int32_t __tgt_rtl_data_submit(int32_t device_id, void *tgt_ptr, void *hst_ptr,
                               int64_t size) {
   assert(DeviceRTL.isValidDeviceId(device_id) && "device_id is invalid");
diff --git a/openmp/libomptarget/plugins/exports b/openmp/libomptarget/plugins/exports
--- a/openmp/libomptarget/plugins/exports
+++ b/openmp/libomptarget/plugins/exports
@@ -7,6 +7,7 @@
     __tgt_rtl_init_device;
     __tgt_rtl_load_binary;
     __tgt_rtl_data_alloc;
+    __tgt_rtl_create_async_info;
     __tgt_rtl_data_submit;
     __tgt_rtl_data_submit_async;
     __tgt_rtl_data_retrieve;
diff --git a/openmp/libomptarget/src/device.h b/openmp/libomptarget/src/device.h
--- a/openmp/libomptarget/src/device.h
+++ b/openmp/libomptarget/src/device.h
@@ -324,6 +324,10 @@
   /// OFFLOAD_SUCCESS/OFFLOAD_FAIL when succeeds/fails.
   int32_t deleteData(void *TgtPtrBegin);
 
+  /// Create async info, OFFLOAD_SUCCESS/OFFLOAD_FAIL when succeeds/fails
+  /// and sets the internal Queue to !nullptr/nullptr for success/fail
+  int32_t createAsyncInfo(AsyncInfoTy &AsyncInfo);
+
   // Data transfer. When AsyncInfo is nullptr, the transfer will be
   // synchronous.
   // Copy data from host to device
diff --git a/openmp/libomptarget/src/device.cpp b/openmp/libomptarget/src/device.cpp
--- a/openmp/libomptarget/src/device.cpp
+++ b/openmp/libomptarget/src/device.cpp
@@ -438,6 +438,12 @@
   return RTL->data_delete(RTLDeviceID, TgtPtrBegin);
 }
 
+int32_t DeviceTy::createAsyncInfo(AsyncInfoTy &AsyncInfo) {
+  if (RTL->create_async_info)
+    return RTL->create_async_info(RTLDeviceID, AsyncInfo);
+  return OFFLOAD_FAIL;
+}
+
 // Submit data to device
 int32_t DeviceTy::submitData(void *TgtPtrBegin, void *HstPtrBegin, int64_t Size,
                              AsyncInfoTy &AsyncInfo) {
@@ -453,7 +459,7 @@
                                 : "unknown");
   }
 
-  if (!AsyncInfo || !RTL->data_submit_async || !RTL->synchronize)
+  if (!RTL->data_submit_async || !RTL->synchronize)
     return RTL->data_submit(RTLDeviceID, TgtPtrBegin, HstPtrBegin, Size);
   else
     return RTL->data_submit_async(RTLDeviceID, TgtPtrBegin, HstPtrBegin, Size,
@@ -484,7 +490,7 @@
 // Copy data from current device to destination device directly
 int32_t DeviceTy::dataExchange(void *SrcPtr, DeviceTy &DstDev, void *DstPtr,
                                int64_t Size, AsyncInfoTy &AsyncInfo) {
-  if (!AsyncInfo || !RTL->data_exchange_async || !RTL->synchronize) {
+  if (!RTL->data_exchange_async || !RTL->synchronize) {
     assert(RTL->data_exchange && "RTL->data_exchange is nullptr");
     return RTL->data_exchange(RTLDeviceID, SrcPtr, DstDev.RTLDeviceID, DstPtr,
                               Size);
@@ -497,7 +503,7 @@
 int32_t DeviceTy::runRegion(void *TgtEntryPtr, void **TgtVarsPtr,
                             ptrdiff_t *TgtOffsets, int32_t TgtVarsSize,
                             AsyncInfoTy &AsyncInfo) {
-  if (!RTL->run_region || !RTL->synchronize)
+  if (!RTL->run_region_async || !RTL->synchronize)
     return RTL->run_region(RTLDeviceID, TgtEntryPtr, TgtVarsPtr, TgtOffsets,
                            TgtVarsSize);
   else
diff --git a/openmp/libomptarget/src/omptarget.cpp b/openmp/libomptarget/src/omptarget.cpp
--- a/openmp/libomptarget/src/omptarget.cpp
+++ b/openmp/libomptarget/src/omptarget.cpp
@@ -19,6 +19,15 @@
 #include <cassert>
 #include <vector>
 
+AsyncInfoTy::AsyncInfoTy(DeviceTy &Device) : Device(Device) {
+  int Result = Device.createAsyncInfo(*this);
+  assert(
+      ((Result == OFFLOAD_SUCCESS) && (AsyncInfo.Queue != nullptr) ||
+       (Result != OFFLOAD_SUCCESS) && (AsyncInfo.Queue == nullptr)) &&
+      "The device plugin returned inconsistent return value and queue pointer");
+  (void)Result;
+}
+
 int AsyncInfoTy::synchronize() {
   int Result = OFFLOAD_SUCCESS;
   if (AsyncInfo.Queue) {
diff --git a/openmp/libomptarget/src/rtl.h b/openmp/libomptarget/src/rtl.h
--- a/openmp/libomptarget/src/rtl.h
+++ b/openmp/libomptarget/src/rtl.h
@@ -31,6 +31,7 @@
   typedef int32_t(init_device_ty)(int32_t);
   typedef __tgt_target_table *(load_binary_ty)(int32_t, void *);
   typedef void *(data_alloc_ty)(int32_t, int64_t, void *, int32_t);
+  typedef int32_t(create_async_info_ty)(int32_t, __tgt_async_info *);
   typedef int32_t(data_submit_ty)(int32_t, void *, void *, int64_t);
   typedef int32_t(data_submit_async_ty)(int32_t, void *, void *, int64_t,
                                         __tgt_async_info *);
@@ -82,6 +83,7 @@
   init_device_ty *init_device = nullptr;
   load_binary_ty *load_binary = nullptr;
   data_alloc_ty *data_alloc = nullptr;
+  create_async_info_ty *create_async_info = nullptr;
   data_submit_ty *data_submit = nullptr;
   data_submit_async_ty *data_submit_async = nullptr;
   data_retrieve_ty *data_retrieve = nullptr;
diff --git a/openmp/libomptarget/src/rtl.cpp b/openmp/libomptarget/src/rtl.cpp
--- a/openmp/libomptarget/src/rtl.cpp
+++ b/openmp/libomptarget/src/rtl.cpp
@@ -154,6 +154,8 @@
     // Optional functions
     *((void **)&R.init_requires) =
         dlsym(dynlib_handle, "__tgt_rtl_init_requires");
+    *((void **)&R.create_async_info) =
+        dlsym(dynlib_handle, "__tgt_rtl_create_async_info");
     *((void **)&R.data_submit_async) =
         dlsym(dynlib_handle, "__tgt_rtl_data_submit_async");
     *((void **)&R.data_retrieve_async) =