diff --git a/examples/android/android-simple/app/src/main/cpp/KomputeModelML.cpp b/examples/android/android-simple/app/src/main/cpp/KomputeModelML.cpp
index e930ddff..b0b7d4df 100644
--- a/examples/android/android-simple/app/src/main/cpp/KomputeModelML.cpp
+++ b/examples/android/android-simple/app/src/main/cpp/KomputeModelML.cpp
@@ -48,13 +48,13 @@ KomputeModelML::train(std::vector<float> yData,
         std::shared_ptr<kp::Algorithm> algorithm = mgr.algorithm(
           params, shader, kp::Workgroup({ 5 }), std::vector<float>({ 5.0 }));
 
-        mgr.sequence()->eval<kp::OpTensorSyncDevice>(params);
+        mgr.sequence()->eval<kp::OpSyncDevice>(params);
 
         std::shared_ptr<kp::Sequence> sq =
           mgr.sequence()
-            ->record<kp::OpTensorSyncDevice>({ wIn, bIn })
+            ->record<kp::OpSyncDevice>({ wIn, bIn })
             ->record<kp::OpAlgoDispatch>(algorithm)
-            ->record<kp::OpTensorSyncLocal>({ wOutI, wOutJ, bOut, lOut });
+            ->record<kp::OpSyncLocal>({ wOutI, wOutJ, bOut, lOut });
 
         // Iterate across all expected iterations
         for (size_t i = 0; i < ITERATIONS; i++) {
diff --git a/examples/array_multiplication/src/main.cpp b/examples/array_multiplication/src/main.cpp
index 825bad7d..96c8267b 100644
--- a/examples/array_multiplication/src/main.cpp
+++ b/examples/array_multiplication/src/main.cpp
@@ -27,9 +27,9 @@ main()
     std::shared_ptr<kp::Algorithm> algo = mgr.algorithm(params, shader);
 
     mgr.sequence()
-      ->record<kp::OpTensorSyncDevice>(params)
+      ->record<kp::OpSyncDevice>(params)
       ->record<kp::OpAlgoDispatch>(algo)
-      ->record<kp::OpTensorSyncLocal>(params)
+      ->record<kp::OpSyncLocal>(params)
       ->eval();
 
     // prints "Output {  0  4  12  }"
diff --git a/examples/logistic_regression/src/main.cpp b/examples/logistic_regression/src/main.cpp
index 0648de08..2d76eed8 100644
--- a/examples/logistic_regression/src/main.cpp
+++ b/examples/logistic_regression/src/main.cpp
@@ -32,7 +32,7 @@ main()
                                                         wIn, wOutI, wOutJ,
                                                         bIn, bOut,  lOut };
 
-    mgr.sequence()->eval<kp::OpTensorSyncDevice>(params);
+    mgr.sequence()->eval<kp::OpSyncDevice>(params);
 
     std::vector<uint32_t> spirv2{ 0x1, 0x2 };
 
@@ -44,9 +44,9 @@ main()
 
     std::shared_ptr<kp::Sequence> sq =
       mgr.sequence()
-        ->record<kp::OpTensorSyncDevice>({ wIn, bIn })
+        ->record<kp::OpSyncDevice>({ wIn, bIn })
         ->record<kp::OpAlgoDispatch>(algorithm)
-        ->record<kp::OpTensorSyncLocal>({ wOutI, wOutJ, bOut, lOut });
+        ->record<kp::OpSyncLocal>({ wOutI, wOutJ, bOut, lOut });
 
     // Iterate across all expected iterations
     for (size_t i = 0; i < ITERATIONS; i++) {
diff --git a/examples/neural_network_vgg7/run_vgg7.py b/examples/neural_network_vgg7/run_vgg7.py
index f5d88e84..67435d69 100644
--- a/examples/neural_network_vgg7/run_vgg7.py
+++ b/examples/neural_network_vgg7/run_vgg7.py
@@ -85,7 +85,7 @@
     if i == 0:
         # For first layer, the input isn't on-device yet
         things_to_sync_to_device.append(tensor_in)
-    last_seq.eval_async(kp.OpTensorSyncDevice(things_to_sync_to_device))
+    last_seq.eval_async(kp.OpSyncDevice(things_to_sync_to_device))
     last_seq.eval_await()
 
     # Prepare
@@ -101,7 +101,7 @@
         # DEBUG:
         # We want to see the output, copy it to local
         last_seq = kpm.sequence()
-        last_seq.eval_async(kp.OpTensorSyncLocal([tensor_out]))
+        last_seq.eval_async(kp.OpSyncLocal([tensor_out]))
         last_seq.eval_await()
         tensor_out.data().astype("<f4", "C").tofile("raw" + str(i) + ".bin")
 
@@ -114,7 +114,7 @@
 
 # Download output
 fin_seq = kpm.sequence()
-fin_seq.eval_async(kp.OpTensorSyncLocal([tensor_in]))
+fin_seq.eval_async(kp.OpSyncLocal([tensor_in]))
 fin_seq.eval_await()
 
 # Output
diff --git a/examples/python_naive_matmul/benchmark.py b/examples/python_naive_matmul/benchmark.py
index 768a854d..51d825c2 100644
--- a/examples/python_naive_matmul/benchmark.py
+++ b/examples/python_naive_matmul/benchmark.py
@@ -31,7 +31,7 @@ def main():
 
         for MatMulOp in [MatMulOp1, MatMulOp2, MatMulOp3]:
             tensor_out.data()[:] = 0
-            mgr.sequence().record(kp.OpTensorSyncDevice([tensor_out]))
+            mgr.sequence().record(kp.OpSyncDevice([tensor_out]))
             matmul_op = MatMulOp(mgr)
             matmul_op(tensor_shape, tensor_in_1, tensor_in_2, tensor_out)
 
diff --git a/examples/python_naive_matmul/first_example.py b/examples/python_naive_matmul/first_example.py
index 1e7caa87..bd20790a 100644
--- a/examples/python_naive_matmul/first_example.py
+++ b/examples/python_naive_matmul/first_example.py
@@ -48,9 +48,9 @@ def main():
         [])  # push_consts
 
     (mgr.sequence()
-     .record(kp.OpTensorSyncDevice(params))
+     .record(kp.OpSyncDevice(params))
      .record(kp.OpAlgoDispatch(algo))
-     .record(kp.OpTensorSyncLocal(params))
+     .record(kp.OpSyncLocal(params))
      .eval())
 
     print(f'Output :\n{tensor_out.data().reshape(tensor_shape)}')
diff --git a/examples/python_naive_matmul/imp1_naive.py b/examples/python_naive_matmul/imp1_naive.py
index a791662d..c65124ca 100644
--- a/examples/python_naive_matmul/imp1_naive.py
+++ b/examples/python_naive_matmul/imp1_naive.py
@@ -90,9 +90,9 @@ def __call__(self, tensor_shape: tuple[int, int], tensor_in_1: kp.Tensor, tensor
                 [])  # push_consts
 
         (self.mgr.sequence()
-         .record(kp.OpTensorSyncDevice([tensor_in_1, tensor_in_2]))
+         .record(kp.OpSyncDevice([tensor_in_1, tensor_in_2]))
          .record(kp.OpAlgoDispatch(self.algo))
-         .record(kp.OpTensorSyncLocal([tensor_out]))
+         .record(kp.OpSyncLocal([tensor_out]))
          .eval())
 
 
diff --git a/examples/python_naive_matmul/imp2_tiled.py b/examples/python_naive_matmul/imp2_tiled.py
index 1ac13e85..4c97c91c 100644
--- a/examples/python_naive_matmul/imp2_tiled.py
+++ b/examples/python_naive_matmul/imp2_tiled.py
@@ -89,9 +89,9 @@ def __call__(self, tensor_shape: tuple[int, int], tensor_in_1: kp.Tensor, tensor
                 [])  # push_consts
 
         (self.mgr.sequence()
-         .record(kp.OpTensorSyncDevice([tensor_in_1, tensor_in_2]))
+         .record(kp.OpSyncDevice([tensor_in_1, tensor_in_2]))
          .record(kp.OpAlgoDispatch(self.algo))
-         .record(kp.OpTensorSyncLocal([tensor_out]))
+         .record(kp.OpSyncLocal([tensor_out]))
          .eval())
 
 
diff --git a/examples/python_naive_matmul/imp3_better_tiling.py b/examples/python_naive_matmul/imp3_better_tiling.py
index 8cd44277..2f2be60e 100644
--- a/examples/python_naive_matmul/imp3_better_tiling.py
+++ b/examples/python_naive_matmul/imp3_better_tiling.py
@@ -110,9 +110,9 @@ def __call__(self, tensor_shape: tuple[int, int], tensor_in_1: kp.Tensor, tensor
                 [])  # push_consts
 
         (self.mgr.sequence()
-         .record(kp.OpTensorSyncDevice([tensor_in_1, tensor_in_2]))
+         .record(kp.OpSyncDevice([tensor_in_1, tensor_in_2]))
          .record(kp.OpAlgoDispatch(self.algo))
-         .record(kp.OpTensorSyncLocal([tensor_out]))
+         .record(kp.OpSyncLocal([tensor_out]))
          .eval())
 
 
diff --git a/python/src/docstrings.hpp b/python/src/docstrings.hpp
index be8e180e..21ed0f15 100644
--- a/python/src/docstrings.hpp
+++ b/python/src/docstrings.hpp
@@ -231,6 +231,8 @@ data storage or transfer.)doc";
 
 static const char *__doc_kp_Image_2 = R"doc()doc";
 
+static const char *__doc_kp_Image_3 = R"doc()doc";
+
 static const char *__doc_kp_ImageT = R"doc()doc";
 
 static const char *__doc_kp_ImageT_ImageT = R"doc()doc";
@@ -399,12 +401,8 @@ static const char *__doc_kp_Image_getPrimaryImage = R"doc()doc";
 
 static const char *__doc_kp_Image_getPrimaryImageUsageFlags = R"doc()doc";
 
-static const char *__doc_kp_Image_getPrimaryMemoryPropertyFlags = R"doc()doc";
-
 static const char *__doc_kp_Image_getStagingImageUsageFlags = R"doc()doc";
 
-static const char *__doc_kp_Image_getStagingMemoryPropertyFlags = R"doc()doc";
-
 static const char *__doc_kp_Image_getWidth =
 R"doc(Retreive the width in the image in pixels
 
@@ -704,12 +702,10 @@ static const char *__doc_kp_Manager_mManageResources = R"doc()doc";
 
 static const char *__doc_kp_Manager_mManagedAlgorithms = R"doc()doc";
 
-static const char *__doc_kp_Manager_mManagedImages = R"doc()doc";
+static const char *__doc_kp_Manager_mManagedMemObjects = R"doc()doc";
 
 static const char *__doc_kp_Manager_mManagedSequences = R"doc()doc";
 
-static const char *__doc_kp_Manager_mManagedTensors = R"doc()doc";
-
 static const char *__doc_kp_Manager_mPhysicalDevice = R"doc()doc";
 
 static const char *__doc_kp_Manager_sequence =
@@ -845,8 +841,14 @@ R"doc(Returns the total size of a single element of the respective data type
     Unsigned integer representing the memory of a single element of
     the respective data type.)doc";
 
+static const char *__doc_kp_Memory_destroy = R"doc(Destroys and frees the GPU resources which includes the memory.)doc";
+
 static const char *__doc_kp_Memory_getDescriptorType = R"doc()doc";
 
+static const char *__doc_kp_Memory_getPrimaryMemoryPropertyFlags = R"doc()doc";
+
+static const char *__doc_kp_Memory_getStagingMemoryPropertyFlags = R"doc()doc";
+
 static const char *__doc_kp_Memory_isInit =
 R"doc(Check whether memory object is initialized based on the created gpu
 resources.
@@ -905,6 +907,21 @@ nullptr when the Tensor is removed.
 Returns:
     Pointer to raw memory containing raw bytes data of Tensor/Image.)doc";
 
+static const char *__doc_kp_Memory_recordCopyFrom =
+R"doc(Records a copy from the memory provided to the current memory. This is
+intended to pass memory into a processing, to perform a staging image
+transfer, or to gather output (between others).
+
+Parameter ``commandBuffer``:
+    Vulkan Command Buffer to record the commands into
+
+Parameter ``copyFromMemory``:
+    Memory to copy the data from)doc";
+
+static const char *__doc_kp_Memory_recordCopyFrom_2 = R"doc()doc";
+
+static const char *__doc_kp_Memory_recordCopyFrom_3 = R"doc()doc";
+
 static const char *__doc_kp_Memory_recordCopyFromDeviceToStaging =
 R"doc(Records a copy from the internal device memory to the staging memory
 using an optional barrier to wait for the operation. This function
@@ -1077,151 +1094,37 @@ submitted as a batch into the GPU.
 Parameter ``commandBuffer``:
     The command buffer to record the command into.)doc";
 
-static const char *__doc_kp_OpImageCopy =
-R"doc(Operation that copies the data from the first image to the rest of the
-images provided, using a record command for all the vectors. This
-operation does not own/manage the memory of the images passed to it.
-The operation must only receive images of type)doc";
-
-static const char *__doc_kp_OpImageCopyToTensor =
-R"doc(Operation that copies the data from the first image to the rest of the
-images provided, using a record command for all the vectors. This
-operation does not own/manage the memory of the images passed to it.
-The operation must only receive images of type)doc";
-
-static const char *__doc_kp_OpImageCopyToTensor_OpImageCopyToTensor =
-R"doc(Default constructor with parameters that provides the core vulkan
-resources and the images that will be used in the operation.
-
-Parameter ``images``:
-    Images that will be used to create in operation.)doc";
-
-static const char *__doc_kp_OpImageCopyToTensor_mImage = R"doc()doc";
-
-static const char *__doc_kp_OpImageCopyToTensor_mTensors = R"doc()doc";
-
-static const char *__doc_kp_OpImageCopyToTensor_postEval =
-R"doc(Copies the local vectors for all the images to sync the data with the
-gpu.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
-
-static const char *__doc_kp_OpImageCopyToTensor_preEval =
-R"doc(Does not perform any preEval commands.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
-
-static const char *__doc_kp_OpImageCopyToTensor_record =
-R"doc(Records the copy commands from the first image into all the other
-images provided. Also optionally records a barrier.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
-
-static const char *__doc_kp_OpImageCopy_OpImageCopy =
-R"doc(Default constructor with parameters that provides the core vulkan
-resources and the images that will be used in the operation.
-
-Parameter ``images``:
-    Images that will be used to create in operation.)doc";
-
-static const char *__doc_kp_OpImageCopy_mImages = R"doc()doc";
-
-static const char *__doc_kp_OpImageCopy_postEval =
-R"doc(Copies the local vectors for all the images to sync the data with the
-gpu.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
-
-static const char *__doc_kp_OpImageCopy_preEval =
-R"doc(Does not perform any preEval commands.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
-
-static const char *__doc_kp_OpImageCopy_record =
-R"doc(Records the copy commands from the first image into all the other
-images provided. Also optionally records a barrier.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
-
-static const char *__doc_kp_OpImageSyncDevice =
-R"doc(Operation that syncs image's device by mapping local data into the
-device memory. For ImageTypes::eDevice it will use a record operation
-for the memory to be syncd into GPU memory which means that the
-operation will be done in sync with GPU commands. For
-ImageTypes::eHost it will only map the data into host memory which
-will happen during preEval before the recorded commands are
-dispatched.)doc";
-
-static const char *__doc_kp_OpImageSyncDevice_OpImageSyncDevice =
-R"doc(Default constructor with parameters that provides the core vulkan
-resources and the images that will be used in the operation. The
-tensos provided cannot be of type ImageTypes::eStorage.
-
-Parameter ``images``:
-    Images that will be used to create in operation.)doc";
-
-static const char *__doc_kp_OpImageSyncDevice_mImages = R"doc()doc";
-
-static const char *__doc_kp_OpImageSyncDevice_postEval =
-R"doc(Does not perform any postEval commands.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
-
-static const char *__doc_kp_OpImageSyncDevice_preEval =
-R"doc(Does not perform any preEval commands.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
-
-static const char *__doc_kp_OpImageSyncDevice_record =
-R"doc(For device images, it records the copy command for the image to copy
-the data from its staging to device memory.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
+static const char *__doc_kp_OpCopy =
+R"doc(Operation that copies the data from the first memory object to the
+rest of the memory objects provided, using a record command for all
+the vectors. This operation does not own/manage the memory of the
+memory objects passed to it.)doc";
 
-static const char *__doc_kp_OpImageSyncLocal =
-R"doc(Operation that syncs image's local memory by mapping device data into
-the local CPU memory. For ImageTypes::eDevice it will use a record
-operation for the memory to be syncd into GPU memory which means that
-the operation will be done in sync with GPU commands. For
-ImageTypes::eHost it will only map the data into host memory which
-will happen during preEval before the recorded commands are
-dispatched.)doc";
-
-static const char *__doc_kp_OpImageSyncLocal_OpImageSyncLocal =
+static const char *__doc_kp_OpCopy_OpCopy =
 R"doc(Default constructor with parameters that provides the core vulkan
-resources and the images that will be used in the operation. The
-images provided cannot be of type ImageTypes::eStorage.
+resources and the memory objects that will be used in the operation.
 
-Parameter ``images``:
-    Images that will be used to create in operation.)doc";
+Parameter ``memObjects``:
+    memory objects that will be used to create in operation.)doc";
 
-static const char *__doc_kp_OpImageSyncLocal_mImages = R"doc()doc";
+static const char *__doc_kp_OpCopy_mMemObjects = R"doc()doc";
 
-static const char *__doc_kp_OpImageSyncLocal_postEval =
-R"doc(For host images it performs the map command from the host memory into
-local memory.
+static const char *__doc_kp_OpCopy_postEval =
+R"doc(Copies the local vectors for all the memory objects to sync the data
+with the gpu.
 
 Parameter ``commandBuffer``:
     The command buffer to record the command into.)doc";
 
-static const char *__doc_kp_OpImageSyncLocal_preEval =
+static const char *__doc_kp_OpCopy_preEval =
 R"doc(Does not perform any preEval commands.
 
 Parameter ``commandBuffer``:
     The command buffer to record the command into.)doc";
 
-static const char *__doc_kp_OpImageSyncLocal_record =
-R"doc(For device images, it records the copy command for the image to copy
-the data from its device to staging memory.
+static const char *__doc_kp_OpCopy_record =
+R"doc(Records the copy commands from the first memory object into all the
+other memory objects provided. Also optionally records a barrier.
 
 Parameter ``commandBuffer``:
     The command buffer to record the command into.)doc";
@@ -1303,117 +1206,45 @@ Parameter ``algorithm``:
     An algorithm that will be overridden with the OpMult shader data
     and the tensors provided which are expected to be 3)doc";
 
-static const char *__doc_kp_OpTensorCopy =
-R"doc(Operation that copies the data from the first tensor to the rest of
-the tensors provided, using a record command for all the vectors. This
-operation does not own/manage the memory of the tensors passed to it.
-The operation must only receive tensors of type)doc";
-
-static const char *__doc_kp_OpTensorCopyToImage =
-R"doc(Operation that copies the data from the first tensor to the vector of
-images provided, using a record command for all the vectors. This
-operation does not own/manage the memory of the tensor/images passed
-to it.)doc";
-
-static const char *__doc_kp_OpTensorCopyToImage_OpTensorCopyToImage =
-R"doc(Default constructor with parameters that provides the core vulkan
-resources and the tensors/images that will be used in the operation.
-
-Parameter ``tensors``:
-    Tensors that will be used to create in operation.)doc";
-
-static const char *__doc_kp_OpTensorCopyToImage_mImages = R"doc()doc";
-
-static const char *__doc_kp_OpTensorCopyToImage_mTensor = R"doc()doc";
-
-static const char *__doc_kp_OpTensorCopyToImage_postEval =
-R"doc(Copies the local vectors for all the tensors to sync the data with the
-gpu.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
-
-static const char *__doc_kp_OpTensorCopyToImage_preEval =
-R"doc(Does not perform any preEval commands.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
-
-static const char *__doc_kp_OpTensorCopyToImage_record =
-R"doc(Records the copy commands from the first tensor into all the other
-tensors provided. Also optionally records a barrier.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
-
-static const char *__doc_kp_OpTensorCopy_OpTensorCopy =
-R"doc(Default constructor with parameters that provides the core vulkan
-resources and the tensors that will be used in the operation.
-
-Parameter ``tensors``:
-    Tensors that will be used to create in operation.)doc";
-
-static const char *__doc_kp_OpTensorCopy_mTensors = R"doc()doc";
-
-static const char *__doc_kp_OpTensorCopy_postEval =
-R"doc(Copies the local vectors for all the tensors to sync the data with the
-gpu.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
-
-static const char *__doc_kp_OpTensorCopy_preEval =
-R"doc(Does not perform any preEval commands.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
-
-static const char *__doc_kp_OpTensorCopy_record =
-R"doc(Records the copy commands from the first tensor into all the other
-tensors provided. Also optionally records a barrier.
-
-Parameter ``commandBuffer``:
-    The command buffer to record the command into.)doc";
-
-static const char *__doc_kp_OpTensorSyncDevice =
-R"doc(Operation that syncs tensor's device by mapping local data into the
-device memory. For TensorTypes::eDevice it will use a record operation
-for the memory to be syncd into GPU memory which means that the
-operation will be done in sync with GPU commands. For
-TensorTypes::eHost it will only map the data into host memory which
+static const char *__doc_kp_OpSyncDevice =
+R"doc(Operation that syncs mem object's device memory by mapping local data
+into the device memory. For MemoryTypes::eDevice it will use a record
+operation for the memory to be syncd into GPU memory which means that
+the operation will be done in sync with GPU commands. For
+MemoryTypes::eHost it will only map the data into host memory which
 will happen during preEval before the recorded commands are
 dispatched.)doc";
 
-static const char *__doc_kp_OpTensorSyncDevice_OpTensorSyncDevice =
+static const char *__doc_kp_OpSyncDevice_OpSyncDevice =
 R"doc(Default constructor with parameters that provides the core vulkan
-resources and the tensors that will be used in the operation. The
-tensos provided cannot be of type TensorTypes::eStorage.
+resources and the memory objects that will be used in the operation.
+The memory objects provided cannot be of type MemoryTypes::eStorage.
 
-Parameter ``tensors``:
-    Tensors that will be used to create in operation.)doc";
+Parameter ``memObjects``:
+    Memory objects that will be used to create in operation.)doc";
 
-static const char *__doc_kp_OpTensorSyncDevice_mTensors = R"doc()doc";
+static const char *__doc_kp_OpSyncDevice_mMemObjects = R"doc()doc";
 
-static const char *__doc_kp_OpTensorSyncDevice_postEval =
+static const char *__doc_kp_OpSyncDevice_postEval =
 R"doc(Does not perform any postEval commands.
 
 Parameter ``commandBuffer``:
     The command buffer to record the command into.)doc";
 
-static const char *__doc_kp_OpTensorSyncDevice_preEval =
+static const char *__doc_kp_OpSyncDevice_preEval =
 R"doc(Does not perform any preEval commands.
 
 Parameter ``commandBuffer``:
     The command buffer to record the command into.)doc";
 
-static const char *__doc_kp_OpTensorSyncDevice_record =
-R"doc(For device tensors, it records the copy command for the tensor to copy
-the data from its staging to device memory.
+static const char *__doc_kp_OpSyncDevice_record =
+R"doc(For device memory objects, it records the copy command for the memory
+object to copy the data from its staging to device memory.
 
 Parameter ``commandBuffer``:
     The command buffer to record the command into.)doc";
 
-static const char *__doc_kp_OpTensorSyncLocal =
+static const char *__doc_kp_OpSyncLocal =
 R"doc(Operation that syncs mem object's local memory by mapping device data
 into the local CPU memory. For MemoryTypes::eDevice it will use a
 record operation for the memory to be syncd into GPU memory which
@@ -1422,32 +1253,32 @@ MemoryTypes::eHost it will only map the data into host memory which
 will happen during preEval before the recorded commands are
 dispatched.)doc";
 
-static const char *__doc_kp_OpTensorSyncLocal_OpTensorSyncLocal =
+static const char *__doc_kp_OpSyncLocal_OpSyncLocal =
 R"doc(Default constructor with parameters that provides the core vulkan
 resources and the memory that will be used in the operation. The
 memory provided cannot be of type MemoryTypes::eStorage.
 
-Parameter ``tensors``:
-    Tensors that will be used to create in operation.)doc";
+Parameter ``memObjects``:
+    Memory objects that will be used to create in operation.)doc";
 
-static const char *__doc_kp_OpTensorSyncLocal_mTensors = R"doc()doc";
+static const char *__doc_kp_OpSyncLocal_mMemObjects = R"doc()doc";
 
-static const char *__doc_kp_OpTensorSyncLocal_postEval =
-R"doc(For host tensors it performs the map command from the host memory into
-local memory.
+static const char *__doc_kp_OpSyncLocal_postEval =
+R"doc(For host memory objects it performs the map command from the host
+memory into local memory.
 
 Parameter ``commandBuffer``:
     The command buffer to record the command into.)doc";
 
-static const char *__doc_kp_OpTensorSyncLocal_preEval =
+static const char *__doc_kp_OpSyncLocal_preEval =
 R"doc(Does not perform any preEval commands.
 
 Parameter ``commandBuffer``:
     The command buffer to record the command into.)doc";
 
-static const char *__doc_kp_OpTensorSyncLocal_record =
-R"doc(For device tensors, it records the copy command for the tensor to copy
-the data from its device to staging memory.
+static const char *__doc_kp_OpSyncLocal_record =
+R"doc(For device memory objects, it records the copy command for the tensor
+to copy the data from its device to staging memory.
 
 Parameter ``commandBuffer``:
     The command buffer to record the command into.)doc";
@@ -1695,7 +1526,9 @@ kp::Algorithms are modified and need to be re-recorded.)doc";
 
 static const char *__doc_kp_Sequence_timestampQueryPool = R"doc()doc";
 
-static const char *__doc_kp_Tensor =
+static const char *__doc_kp_Tensor = R"doc()doc";
+
+static const char *__doc_kp_Tensor_2 =
 R"doc(Structured data used in GPU operations.
 
 Tensors are the base building block in Kompute to perform operations
@@ -1778,12 +1611,8 @@ static const char *__doc_kp_Tensor_getPrimaryBuffer = R"doc()doc";
 
 static const char *__doc_kp_Tensor_getPrimaryBufferUsageFlags = R"doc()doc";
 
-static const char *__doc_kp_Tensor_getPrimaryMemoryPropertyFlags = R"doc()doc";
-
 static const char *__doc_kp_Tensor_getStagingBufferUsageFlags = R"doc()doc";
 
-static const char *__doc_kp_Tensor_getStagingMemoryPropertyFlags = R"doc()doc";
-
 static const char *__doc_kp_Tensor_isInit =
 R"doc(Check whether tensor is initialized based on the created gpu
 resources.
diff --git a/python/src/main.cpp b/python/src/main.cpp
index 4d12d3ec..210fbe04 100644
--- a/python/src/main.cpp
+++ b/python/src/main.cpp
@@ -80,42 +80,24 @@ PYBIND11_MODULE(kp, m)
     py::class_<kp::OpBase, std::shared_ptr<kp::OpBase>>(
       m, "OpBase", DOC(kp, OpBase));
 
-    py::class_<kp::OpTensorSyncDevice,
+    py::class_<kp::OpSyncDevice,
                kp::OpBase,
-               std::shared_ptr<kp::OpTensorSyncDevice>>(
-      m, "OpTensorSyncDevice", DOC(kp, OpTensorSyncDevice))
+               std::shared_ptr<kp::OpSyncDevice>>(
+      m, "OpSyncDevice", DOC(kp, OpSyncDevice))
       .def(py::init<const std::vector<std::shared_ptr<kp::Memory>>&>(),
-           DOC(kp, OpTensorSyncDevice, OpTensorSyncDevice));
+           DOC(kp, OpSyncDevice, OpSyncDevice));
 
-    py::class_<kp::OpTensorSyncLocal,
+    py::class_<kp::OpSyncLocal,
                kp::OpBase,
-               std::shared_ptr<kp::OpTensorSyncLocal>>(
-      m, "OpTensorSyncLocal", DOC(kp, OpTensorSyncLocal))
+               std::shared_ptr<kp::OpSyncLocal>>(
+      m, "OpSyncLocal", DOC(kp, OpSyncLocal))
       .def(py::init<const std::vector<std::shared_ptr<kp::Memory>>&>(),
-           DOC(kp, OpTensorSyncLocal, OpTensorSyncLocal));
+           DOC(kp, OpSyncLocal, OpSyncLocal));
 
-    py::class_<kp::OpTensorCopy, kp::OpBase, std::shared_ptr<kp::OpTensorCopy>>(
-      m, "OpTensorCopy", DOC(kp, OpTensorCopy))
+    py::class_<kp::OpCopy, kp::OpBase, std::shared_ptr<kp::OpCopy>>(
+      m, "OpCopy", DOC(kp, OpCopy))
       .def(py::init<const std::vector<std::shared_ptr<kp::Memory>>&>(),
-           DOC(kp, OpTensorCopy, OpTensorCopy));
-    py::class_<kp::OpImageSyncDevice,
-               kp::OpBase,
-               std::shared_ptr<kp::OpImageSyncDevice>>(
-      m, "OpImageSyncDevice", DOC(kp, OpImageSyncDevice))
-      .def(py::init<const std::vector<std::shared_ptr<kp::Memory>>&>(),
-           DOC(kp, OpImageSyncDevice, OpImageSyncDevice));
-
-    py::class_<kp::OpImageSyncLocal,
-               kp::OpBase,
-               std::shared_ptr<kp::OpImageSyncLocal>>(
-      m, "OpImageSyncLocal", DOC(kp, OpImageSyncLocal))
-      .def(py::init<const std::vector<std::shared_ptr<kp::Memory>>&>(),
-           DOC(kp, OpImageSyncLocal, OpImageSyncLocal));
-
-    py::class_<kp::OpImageCopy, kp::OpBase, std::shared_ptr<kp::OpImageCopy>>(
-      m, "OpImageCopy", DOC(kp, OpImageCopy))
-      .def(py::init<const std::vector<std::shared_ptr<kp::Memory>>&>(),
-           DOC(kp, OpImageCopy, OpImageCopy));
+           DOC(kp, OpCopy, OpCopy));
 
     py::class_<kp::OpAlgoDispatch,
                kp::OpBase,
diff --git a/python/test/test_array_multiplication.py b/python/test/test_array_multiplication.py
index e8de6832..b8cc498f 100644
--- a/python/test/test_array_multiplication.py
+++ b/python/test/test_array_multiplication.py
@@ -25,9 +25,9 @@ def compute_mult(index=("input", "GlobalInvocationId", ps.ivec3),
         data3[i] = data1[i] * data2[i]
 
     (mgr.sequence()
-        .record(kp.OpTensorSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(mgr.algorithm(params, compute_mult.to_spirv())))
-        .record(kp.OpTensorSyncLocal([tensor_out]))
+        .record(kp.OpSyncLocal([tensor_out]))
         .eval())
 
     assert tensor_out.data().tolist() == [2.0, 4.0, 6.0]
diff --git a/python/test/test_image_types.py b/python/test/test_image_types.py
index f0f89a6c..39d93714 100644
--- a/python/test/test_image_types.py
+++ b/python/test/test_image_types.py
@@ -36,9 +36,9 @@ def test_type_float():
     params = [image_in_a, image_in_b, image_out]
 
     (mgr.sequence()
-        .record(kp.OpImageSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(mgr.algorithm(params, spirv)))
-        .record(kp.OpImageSyncLocal([image_out]))
+        .record(kp.OpSyncLocal([image_out]))
         .eval())
 
     assert np.all(image_out.data() == arr_in_a * arr_in_b)
@@ -74,9 +74,9 @@ def test_type_int():
     params = [image_in_a, image_in_b, image_out]
 
     (mgr.sequence()
-        .record(kp.OpImageSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(mgr.algorithm(params, spirv)))
-        .record(kp.OpImageSyncLocal([image_out]))
+        .record(kp.OpSyncLocal([image_out]))
         .eval())
 
     print(f"Dtype value {image_out.data().dtype}")
@@ -114,9 +114,9 @@ def test_type_unsgined_int():
     params = [image_in_a, image_in_b, image_out]
 
     (mgr.sequence()
-        .record(kp.OpImageSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(mgr.algorithm(params, spirv)))
-        .record(kp.OpImageSyncLocal([image_out]))
+        .record(kp.OpSyncLocal([image_out]))
         .eval())
 
     print(f"Dtype value {image_out.data().dtype}")
@@ -154,9 +154,9 @@ def test_type_short():
     params = [image_in_a, image_in_b, image_out]
 
     (mgr.sequence()
-        .record(kp.OpImageSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(mgr.algorithm(params, spirv)))
-        .record(kp.OpImageSyncLocal([image_out]))
+        .record(kp.OpSyncLocal([image_out]))
         .eval())
 
     print(f"Dtype value {image_out.data().dtype}")
@@ -194,9 +194,9 @@ def test_type_unsgined_short():
     params = [image_in_a, image_in_b, image_out]
 
     (mgr.sequence()
-        .record(kp.OpImageSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(mgr.algorithm(params, spirv)))
-        .record(kp.OpImageSyncLocal([image_out]))
+        .record(kp.OpSyncLocal([image_out]))
         .eval())
 
     print(f"Dtype value {image_out.data().dtype}")
@@ -234,9 +234,9 @@ def test_type_char():
     params = [image_in_a, image_in_b, image_out]
 
     (mgr.sequence()
-        .record(kp.OpImageSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(mgr.algorithm(params, spirv)))
-        .record(kp.OpImageSyncLocal([image_out]))
+        .record(kp.OpSyncLocal([image_out]))
         .eval())
 
     print(f"Dtype value {image_out.data().dtype}")
@@ -274,9 +274,9 @@ def test_type_unsgined_char():
     params = [image_in_a, image_in_b, image_out]
 
     (mgr.sequence()
-        .record(kp.OpImageSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(mgr.algorithm(params, spirv)))
-        .record(kp.OpImageSyncLocal([image_out]))
+        .record(kp.OpSyncLocal([image_out]))
         .eval())
 
     print(f"Dtype value {image_out.data().dtype}")
diff --git a/python/test/test_kompute.py b/python/test/test_kompute.py
index 4b7ec545..c05cca74 100644
--- a/python/test/test_kompute.py
+++ b/python/test/test_kompute.py
@@ -58,13 +58,13 @@ def test_end_to_end():
     algo = mgr.algorithm(params, compile_source(shader), workgroup, spec_consts, push_consts_a)
 
     (mgr.sequence()
-        .record(kp.OpTensorSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(algo))
         .record(kp.OpAlgoDispatch(algo, push_consts_b))
         .eval())
 
     sq = mgr.sequence()
-    sq.eval_async(kp.OpTensorSyncLocal(params))
+    sq.eval_async(kp.OpSyncLocal(params))
 
     sq.eval_await()
 
@@ -104,9 +104,9 @@ def test_shader_str():
     algo = mgr.algorithm(params, spirv)
 
     (mgr.sequence()
-        .record(kp.OpTensorSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(algo))
-        .record(kp.OpTensorSyncLocal(params))
+        .record(kp.OpSyncLocal(params))
         .eval())
 
     assert tensor_out.data().tolist() == [2.0, 4.0, 6.0]
@@ -145,9 +145,9 @@ def test_sequence():
 
     sq = mgr.sequence()
 
-    sq.record(kp.OpTensorSyncDevice(params))
+    sq.record(kp.OpSyncDevice(params))
     sq.record(kp.OpAlgoDispatch(algo))
-    sq.record(kp.OpTensorSyncLocal(params))
+    sq.record(kp.OpSyncLocal(params))
 
     sq.eval()
 
@@ -194,11 +194,11 @@ def test_pushconsts():
     algo = mgr.algorithm([tensor], spirv, (1, 1, 1), [], [0.1, 0.2, 0.3])
 
     (mgr.sequence()
-        .record(kp.OpTensorSyncDevice([tensor]))
+        .record(kp.OpSyncDevice([tensor]))
         .record(kp.OpAlgoDispatch(algo))
         .record(kp.OpAlgoDispatch(algo, [0.3, 0.2, 0.1]))
         .record(kp.OpAlgoDispatch(algo, [0.3, 0.2, 0.1]))
-        .record(kp.OpTensorSyncLocal([tensor]))
+        .record(kp.OpSyncLocal([tensor]))
         .eval())
 
     assert np.allclose(tensor.data(), np.array([0.7, 0.6, 0.5], dtype=np.float32))
@@ -232,11 +232,11 @@ def test_pushconsts_int():
     algo = mgr.algorithm([tensor], spirv, (1, 1, 1), spec_consts, push_consts)
 
     (mgr.sequence()
-        .record(kp.OpTensorSyncDevice([tensor]))
+        .record(kp.OpSyncDevice([tensor]))
         .record(kp.OpAlgoDispatch(algo))
         .record(kp.OpAlgoDispatch(algo, np.array([-1, -1, -1], dtype=np.int32)))
         .record(kp.OpAlgoDispatch(algo, np.array([-1, -1, -1], dtype=np.int32)))
-        .record(kp.OpTensorSyncLocal([tensor]))
+        .record(kp.OpSyncLocal([tensor]))
         .eval())
 
     assert np.all(tensor.data() == np.array([-3, -3, -3], dtype=np.int32))
@@ -261,9 +261,9 @@ def compute_shader_wg(gl_idx=("input", "GlobalInvocationId", ps.ivec3),
     algo = mgr.algorithm([tensor_a, tensor_b], compute_shader_wg.to_spirv(), (16,8,1))
 
     (mgr.sequence()
-        .record(kp.OpTensorSyncDevice([tensor_a, tensor_b]))
+        .record(kp.OpSyncDevice([tensor_a, tensor_b]))
         .record(kp.OpAlgoDispatch(algo))
-        .record(kp.OpTensorSyncLocal([tensor_a, tensor_b]))
+        .record(kp.OpSyncLocal([tensor_a, tensor_b]))
         .eval())
 
     print(tensor_a.data())
diff --git a/python/test/test_logistic_regression.py b/python/test/test_logistic_regression.py
index f7c50aa7..12eb910d 100644
--- a/python/test/test_logistic_regression.py
+++ b/python/test/test_logistic_regression.py
@@ -67,19 +67,19 @@ def compute_shader(
     params = [tensor_x_i, tensor_x_j, tensor_y, tensor_w_in, tensor_w_out_i,
         tensor_w_out_j, tensor_b_in, tensor_b_out, tensor_l_out, tensor_m]
 
-    mgr.sequence().eval(kp.OpTensorSyncDevice(params))
+    mgr.sequence().eval(kp.OpSyncDevice(params))
 
     # Create a managed sequence
     sq = mgr.sequence()
 
     # Record operation to sync memory from local to GPU memory
-    sq.record(kp.OpTensorSyncDevice([tensor_w_in, tensor_b_in]))
+    sq.record(kp.OpSyncDevice([tensor_w_in, tensor_b_in]))
 
     # Record operation to execute GPU shader against all our parameters
     sq.record(kp.OpAlgoDispatch(mgr.algorithm(params, compute_shader.to_spirv())))
 
     # Record operation to sync memory from GPU to local memory
-    sq.record(kp.OpTensorSyncLocal([tensor_w_out_i, tensor_w_out_j, tensor_b_out, tensor_l_out]))
+    sq.record(kp.OpSyncLocal([tensor_w_out_i, tensor_w_out_j, tensor_b_out, tensor_l_out]))
 
     ITERATIONS = 100
     learning_rate = 0.1
diff --git a/python/test/test_tensor_types.py b/python/test/test_tensor_types.py
index d615a55b..1f8b603e 100644
--- a/python/test/test_tensor_types.py
+++ b/python/test/test_tensor_types.py
@@ -38,9 +38,9 @@ def test_type_float():
     params = [tensor_in_a, tensor_in_b, tensor_out]
 
     (mgr.sequence()
-        .record(kp.OpTensorSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(mgr.algorithm(params, spirv)))
-        .record(kp.OpTensorSyncLocal([tensor_out]))
+        .record(kp.OpSyncLocal([tensor_out]))
         .eval())
 
     assert np.all(tensor_out.data() == arr_in_a * arr_in_b)
@@ -77,9 +77,9 @@ def test_type_float_double_incorrect():
     params = [tensor_in_a, tensor_in_b, tensor_out]
 
     (mgr.sequence()
-        .record(kp.OpTensorSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(mgr.algorithm(params, spirv)))
-        .record(kp.OpTensorSyncLocal([tensor_out]))
+        .record(kp.OpSyncLocal([tensor_out]))
         .eval())
 
     assert np.all(tensor_out.data() != arr_in_a * arr_in_b)
@@ -119,9 +119,9 @@ def test_type_double():
     params = [tensor_in_a, tensor_in_b, tensor_out]
 
     (mgr.sequence()
-        .record(kp.OpTensorSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(mgr.algorithm(params, spirv)))
-        .record(kp.OpTensorSyncLocal([tensor_out]))
+        .record(kp.OpSyncLocal([tensor_out]))
         .eval())
 
     print(f"Dtype value {tensor_out.data().dtype}")
@@ -159,9 +159,9 @@ def test_type_int():
     params = [tensor_in_a, tensor_in_b, tensor_out]
 
     (mgr.sequence()
-        .record(kp.OpTensorSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(mgr.algorithm(params, spirv)))
-        .record(kp.OpTensorSyncLocal([tensor_out]))
+        .record(kp.OpSyncLocal([tensor_out]))
         .eval())
 
     print(f"Dtype value {tensor_out.data().dtype}")
@@ -199,9 +199,9 @@ def test_type_unsigned_int():
     params = [tensor_in_a, tensor_in_b, tensor_out]
 
     (mgr.sequence()
-        .record(kp.OpTensorSyncDevice(params))
+        .record(kp.OpSyncDevice(params))
         .record(kp.OpAlgoDispatch(mgr.algorithm(params, spirv)))
-        .record(kp.OpTensorSyncLocal([tensor_out]))
+        .record(kp.OpSyncLocal([tensor_out]))
         .eval())
 
     print(f"Dtype value {tensor_out.data().dtype}")