diff --git a/runtime/command_stream/aub_command_stream_receiver_hw.inl b/runtime/command_stream/aub_command_stream_receiver_hw.inl
index 9edc183921..4666d4295a 100644
--- a/runtime/command_stream/aub_command_stream_receiver_hw.inl
+++ b/runtime/command_stream/aub_command_stream_receiver_hw.inl
@@ -378,6 +378,7 @@ FlushStamp AUBCommandStreamReceiverHw<GfxFamily>::flush(BatchBuffer &batchBuffer
 
     if (this->standalone) {
         pollForCompletion(engineType);
+        *this->tagAddress = this->peekLatestSentTaskCount();
     }
     return 0;
 }
diff --git a/runtime/command_stream/create_command_stream_impl.cpp b/runtime/command_stream/create_command_stream_impl.cpp
index 50e0f320fa..4d5f93a04a 100644
--- a/runtime/command_stream/create_command_stream_impl.cpp
+++ b/runtime/command_stream/create_command_stream_impl.cpp
@@ -46,7 +46,6 @@ CommandStreamReceiver *createCommandStreamImpl(const HardwareInfo *pHwInfo) {
                              pHwInfo->pWaTable,
                              pHwInfo->pSysInfo);
             commandStreamReceiver = AUBCommandStreamReceiver::create(*pHwInfo, "aubfile", true);
-            initialHardwareTag = -1;
             break;
         case CSR_TBX:
             Gmm::initContext(pHwInfo->pPlatform,
diff --git a/unit_tests/command_stream/aub_command_stream_receiver_tests.cpp b/unit_tests/command_stream/aub_command_stream_receiver_tests.cpp
index 37cd2418d5..916c82ebe1 100644
--- a/unit_tests/command_stream/aub_command_stream_receiver_tests.cpp
+++ b/unit_tests/command_stream/aub_command_stream_receiver_tests.cpp
@@ -55,6 +55,14 @@ struct MockAubCsr : public AUBCommandStreamReceiverHw<GfxFamily> {
         return this->dispatchMode;
     }
 
+    GraphicsAllocation *getTagAllocation() const {
+        return this->tagAllocation;
+    }
+
+    void setLatestSentTaskCount(uint32_t latestSentTaskCount) {
+        this->latestSentTaskCount = latestSentTaskCount;
+    }
+
     MOCK_METHOD2(flattenBatchBuffer, void *(BatchBuffer &batchBuffer, size_t &sizeBatchBuffer));
 };
 
@@ -125,9 +133,12 @@ HWTEST_F(AubCommandStreamReceiverTests, givenGraphicsAllocationWhenMakeResidentC
 
 HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverWhenFlushIsCalledThenItShouldLeaveProperRingTailAlignment) {
     std::unique_ptr<MemoryManager> memoryManager(nullptr);
-    std::unique_ptr<AUBCommandStreamReceiverHw<FamilyType>> aubCsr(new AUBCommandStreamReceiverHw<FamilyType>(*platformDevices[0], true));
+    std::unique_ptr<MockAubCsr<FamilyType>> aubCsr(new MockAubCsr<FamilyType>(*platformDevices[0], true));
     memoryManager.reset(aubCsr->createMemoryManager(false));
 
+    aubCsr->setTagAllocation(pDevice->getTagAllocation());
+    ASSERT_NE(nullptr, aubCsr->getTagAllocation());
+
     GraphicsAllocation *commandBuffer = memoryManager->allocateGraphicsMemory(4096, 4096);
     ASSERT_NE(nullptr, commandBuffer);
     LinearStream cs(commandBuffer);
@@ -150,11 +161,42 @@ HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverWhenFlushIs
     memoryManager->freeGraphicsMemory(commandBuffer);
 }
 
+HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverInStandaloneModeWhenFlushIsCalledThenItShouldUpdateHwTagWithLatestSentTaskCount) {
+    std::unique_ptr<MemoryManager> memoryManager(nullptr);
+    std::unique_ptr<MockAubCsr<FamilyType>> aubCsr(new MockAubCsr<FamilyType>(*platformDevices[0], true));
+    memoryManager.reset(aubCsr->createMemoryManager(false));
+
+    auto commandBuffer = memoryManager->allocateGraphicsMemory(4096, 4096);
+    ASSERT_NE(nullptr, commandBuffer);
+    LinearStream cs(commandBuffer);
+
+    BatchBuffer batchBuffer{cs.getGraphicsAllocation(), 0, 0, nullptr, false, false, QueueThrottle::MEDIUM, cs.getUsed(), &cs};
+    auto engineType = OCLRT::ENGINE_RCS;
+    ResidencyContainer allocationsForResidency = {};
+
+    aubCsr->setTagAllocation(pDevice->getTagAllocation());
+    ASSERT_NE(nullptr, aubCsr->getTagAllocation());
+    EXPECT_EQ(initialHardwareTag, *aubCsr->getTagAddress());
+
+    aubCsr->setLatestSentTaskCount(aubCsr->peekTaskCount() + 1);
+
+    EXPECT_NE(aubCsr->peekLatestSentTaskCount(), *aubCsr->getTagAddress());
+
+    aubCsr->flush(batchBuffer, engineType, &allocationsForResidency);
+
+    EXPECT_EQ(aubCsr->peekLatestSentTaskCount(), *aubCsr->getTagAddress());
+
+    memoryManager->freeGraphicsMemory(commandBuffer);
+}
+
 HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverInStandaloneModeWhenFlushIsCalledThenItShouldCallMakeResidentOnCommandBufferAllocation) {
     std::unique_ptr<MemoryManager> memoryManager(nullptr);
-    std::unique_ptr<AUBCommandStreamReceiverHw<FamilyType>> aubCsr(new AUBCommandStreamReceiverHw<FamilyType>(*platformDevices[0], true));
+    std::unique_ptr<MockAubCsr<FamilyType>> aubCsr(new MockAubCsr<FamilyType>(*platformDevices[0], true));
     memoryManager.reset(aubCsr->createMemoryManager(false));
 
+    aubCsr->setTagAllocation(pDevice->getTagAllocation());
+    ASSERT_NE(nullptr, aubCsr->getTagAllocation());
+
     GraphicsAllocation *commandBuffer = memoryManager->allocateGraphicsMemory(4096, 4096);
     ASSERT_NE(nullptr, commandBuffer);
     LinearStream cs(commandBuffer);
@@ -201,10 +243,14 @@ HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverInNoneStand
 
 HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverInStandaloneModeWhenFlushIsCalledThenItShouldCallMakeResidentOnResidencyAllocations) {
     std::unique_ptr<MemoryManager> memoryManager(nullptr);
-    std::unique_ptr<AUBCommandStreamReceiverHw<FamilyType>> aubCsr(new AUBCommandStreamReceiverHw<FamilyType>(*platformDevices[0], true));
+    std::unique_ptr<MockAubCsr<FamilyType>> aubCsr(new MockAubCsr<FamilyType>(*platformDevices[0], true));
     memoryManager.reset(aubCsr->createMemoryManager(false));
 
+    aubCsr->setTagAllocation(pDevice->getTagAllocation());
+    ASSERT_NE(nullptr, aubCsr->getTagAllocation());
+
     auto gfxAllocation = memoryManager->allocateGraphicsMemory(sizeof(uint32_t), sizeof(uint32_t), false, false);
+    ASSERT_NE(nullptr, gfxAllocation);
 
     GraphicsAllocation *commandBuffer = memoryManager->allocateGraphicsMemory(4096, 4096);
     ASSERT_NE(nullptr, commandBuffer);
@@ -387,11 +433,14 @@ HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverWhenForcedB
     memoryManager->freeGraphicsMemory(commandBuffer);
 }
 
-HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverWhenDefaultDebugConfigThenExpectxpectFlattenBatchBufferIsNotCalled) {
+HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverWhenDefaultDebugConfigThenExpectFlattenBatchBufferIsNotCalled) {
     std::unique_ptr<MemoryManager> memoryManager(nullptr);
     std::unique_ptr<MockAubCsr<FamilyType>> aubCsr(new MockAubCsr<FamilyType>(*platformDevices[0], true));
     memoryManager.reset(aubCsr->createMemoryManager(false));
 
+    aubCsr->setTagAllocation(pDevice->getTagAllocation());
+    ASSERT_NE(nullptr, aubCsr->getTagAllocation());
+
     GraphicsAllocation *commandBuffer = memoryManager->allocateGraphicsMemory(4096, 4096);
     ASSERT_NE(nullptr, commandBuffer);
     LinearStream cs(commandBuffer);
@@ -415,6 +464,9 @@ HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverWhenForcedF
     std::unique_ptr<MockAubCsr<FamilyType>> aubCsr(new MockAubCsr<FamilyType>(*platformDevices[0], true));
     memoryManager.reset(aubCsr->createMemoryManager(false));
 
+    aubCsr->setTagAllocation(pDevice->getTagAllocation());
+    ASSERT_NE(nullptr, aubCsr->getTagAllocation());
+
     GraphicsAllocation *commandBuffer = memoryManager->allocateGraphicsMemory(4096, 4096);
     ASSERT_NE(nullptr, commandBuffer);
     LinearStream cs(commandBuffer);
@@ -436,7 +488,7 @@ HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverWhenForcedF
     memoryManager->freeGraphicsMemory(chainedBatchBuffer);
 }
 
-HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverWhenForcedFlattenBatchBufferAndImmediateDispatchModeAndTheresNoChainedBatchBufferThenExpectFlattenBatchBufferIsCalledAnyway) {
+HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverWhenForcedFlattenBatchBufferAndImmediateDispatchModeAndThereIsNoChainedBatchBufferThenExpectFlattenBatchBufferIsCalledAnyway) {
     DebugManagerStateRestore dbgRestore;
     DebugManager.flags.FlattenBatchBufferForAUBDump.set(true);
     DebugManager.flags.CsrDispatchMode.set(CommandStreamReceiver::DispatchMode::ImmediateDispatch);
@@ -445,6 +497,9 @@ HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverWhenForcedF
     std::unique_ptr<MockAubCsr<FamilyType>> aubCsr(new MockAubCsr<FamilyType>(*platformDevices[0], true));
     memoryManager.reset(aubCsr->createMemoryManager(false));
 
+    aubCsr->setTagAllocation(pDevice->getTagAllocation());
+    ASSERT_NE(nullptr, aubCsr->getTagAllocation());
+
     GraphicsAllocation *commandBuffer = memoryManager->allocateGraphicsMemory(4096, 4096);
     ASSERT_NE(nullptr, commandBuffer);
     LinearStream cs(commandBuffer);
@@ -467,6 +522,9 @@ HWTEST_F(AubCommandStreamReceiverTests, givenAubCommandStreamReceiverWhenDispatc
     memoryManager.reset(aubCsr->createMemoryManager(false));
     aubCsr->overrideDispatchPolicy(CommandStreamReceiver::DispatchMode::BatchedDispatch);
 
+    aubCsr->setTagAllocation(pDevice->getTagAllocation());
+    ASSERT_NE(nullptr, aubCsr->getTagAllocation());
+
     GraphicsAllocation *commandBuffer = memoryManager->allocateGraphicsMemory(4096, 4096);
     ASSERT_NE(nullptr, commandBuffer);
     LinearStream cs(commandBuffer);