diff --git a/src/OpAlgoDispatch.cpp b/src/OpAlgoDispatch.cpp
index 94502cc76..517a70d52 100644
--- a/src/OpAlgoDispatch.cpp
+++ b/src/OpAlgoDispatch.cpp
@@ -28,9 +28,9 @@ OpAlgoDispatch::record(const vk::CommandBuffer& commandBuffer)
          this->mAlgorithm->getTensors()) {
         tensor->recordPrimaryBufferMemoryBarrier(
           commandBuffer,
-          vk::AccessFlagBits::eHostWrite,
+          vk::AccessFlagBits::eTransferWrite,
           vk::AccessFlagBits::eShaderRead,
-          vk::PipelineStageFlagBits::eHost,
+          vk::PipelineStageFlagBits::eTransfer,
           vk::PipelineStageFlagBits::eComputeShader);
     }
 
diff --git a/src/OpTensorSyncLocal.cpp b/src/OpTensorSyncLocal.cpp
index 70642b9de..8a99c780b 100644
--- a/src/OpTensorSyncLocal.cpp
+++ b/src/OpTensorSyncLocal.cpp
@@ -31,13 +31,13 @@ OpTensorSyncLocal::record(const vk::CommandBuffer& commandBuffer)
     for (size_t i = 0; i < this->mTensors.size(); i++) {
         if (this->mTensors[i]->tensorType() == Tensor::TensorTypes::eDevice) {
 
-            this->mTensors[i]->recordCopyFromDeviceToStaging(commandBuffer);
+            this->mTensors[i]->recordPrimaryBufferMemoryBarrier(commandBuffer,
+                                        vk::AccessFlagBits::eShaderWrite,
+                                        vk::AccessFlagBits::eTransferRead,
+                                        vk::PipelineStageFlagBits::eComputeShader,
+                                        vk::PipelineStageFlagBits::eTransfer);
 
-            this->mTensors[i]->recordStagingBufferMemoryBarrier(commandBuffer,
-                                        vk::AccessFlagBits::eTransferWrite,
-                                        vk::AccessFlagBits::eHostRead,
-                                        vk::PipelineStageFlagBits::eTransfer,
-                                        vk::PipelineStageFlagBits::eHost);
+            this->mTensors[i]->recordCopyFromDeviceToStaging(commandBuffer);
         }
     }
 }