diff --git a/paddle/phi/kernels/sparse/gpu/convolution_grad_kernel.cu b/paddle/phi/kernels/sparse/gpu/convolution_grad_kernel.cu
index e54e39f5541d55a0613554ddfad9cb8a26ca23ff..c19bf67be2611d1ec5d1f00c89df3526ff5ec38c 100644
--- a/paddle/phi/kernels/sparse/gpu/convolution_grad_kernel.cu
+++ b/paddle/phi/kernels/sparse/gpu/convolution_grad_kernel.cu
@@ -22,7 +22,7 @@ limitations under the License. */
 #include "paddle/phi/kernels/copy_kernel.h"
 #include "paddle/phi/kernels/funcs/blas/blas.h"
 #include "paddle/phi/kernels/funcs/math_function.h"
-#include "paddle/phi/kernels/funcs/sparse/scatter.cu.h"
+#include "paddle/phi/kernels/funcs/scatter.cu.h"
 #include "paddle/phi/kernels/sparse/convolution_grad_kernel.h"
 #include "paddle/phi/kernels/sparse/gpu/convolution.cu.h"
 
@@ -203,38 +203,19 @@ void Conv3dGradGPUKernel(const GPUContext& dev_ctx,
   }
 
   // 4. scatter
-  // x_grad->ResizeAndAllocate(x.non_zero_elements().dims());
-  DenseTensorMeta index_meta(DataType::INT32, {rulebook_len}, DataLayout::NCHW);
-  DenseTensor out_index = phi::Empty(dev_ctx, std::move(index_meta));
-  DenseTensor unique_key = phi::Empty(
-      dev_ctx,
-      DenseTensorMeta(paddle::experimental::CppTypeToDataType<IntT>::Type(),
-                      {rulebook_len},
-                      DataLayout::NCHW));
-  DenseTensor unique_value = phi::Empty(dev_ctx, std::move(index_meta));
-
-  SortedAndUniqueIndex<GPUContext, IntT>(dev_ctx,
-                                         rulebook_ptr + rulebook_len,
-                                         rulebook_len,
-                                         &out_index,
-                                         &unique_key,
-                                         &unique_value);
-
   config = phi::backends::gpu::GetGpuLaunchConfig1D(
       dev_ctx, rulebook_len * in_channels, 1);
 
-  phi::funcs::sparse::ScatterKernel<T><<<config.block_per_grid.x,
-                                         config.thread_per_block.x,
-                                         0,
-                                         dev_ctx.stream()>>>(
+  phi::funcs::ScatterCUDAKernel<<<config.block_per_grid,
+                                  config.thread_per_block,
+                                  0,
+                                  dev_ctx.stream()>>>(
       d_x_features_ptr,
-      unique_value.data<int>(),
-      out_index.data<int>(),
-      x.nnz(),
+      rulebook_ptr + rulebook_len,
+      x_grad_values_ptr,
       rulebook_len,
       in_channels,
-      x_grad_values_ptr,
-      subm);
+      false);
 }
 
 template <typename T, typename Context>