diff --git a/lite/kernels/arm/elementwise_compute.cc b/lite/kernels/arm/elementwise_compute.cc
index 92512017348a556fdead7ef8aa02a65023844c11..71fedc966556aadc40848fff54bae7884cd0e9f8 100644
--- a/lite/kernels/arm/elementwise_compute.cc
+++ b/lite/kernels/arm/elementwise_compute.cc
@@ -132,24 +132,25 @@ void ElementwiseAddActivationCompute::Run() {
   }
 }
 
-void ElementwiseSubCompute::Run() {
-  auto& param = Param<operators::ElementwiseParam>();
-  const float* x_data = param.X->data<float>();
-  const float* y_data = param.Y->data<float>();
-  float* out_data = param.Out->mutable_data<float>();
+template <typename T, PrecisionType PType>
+void ElementwiseSubCompute<T, PType>::Run() {
+  auto& param = this->template Param<operators::ElementwiseParam>();
+  const T* x_data = param.X->template data<T>();
+  const T* y_data = param.Y->template data<T>();
+  T* out_data = param.Out->template mutable_data<T>();
   int axis = param.axis;
   auto x_dims = param.X->dims();
   auto y_dims = param.Y->dims();
   int pre, n, post;
   if (x_dims.size() < y_dims.size() &&
       is_broadcast(y_dims, x_dims, axis, &pre, &n, &post)) {
-    lite::arm::math::elementwise_sub_broadcast(
+    lite::arm::math::elementwise_sub_broadcast<T>(
         y_data, x_data, out_data, pre, n, post);
   } else if (is_broadcast(x_dims, y_dims, axis, &pre, &n, &post)) {
-    lite::arm::math::elementwise_sub_broadcast(
+    lite::arm::math::elementwise_sub_broadcast<T>(
         x_data, y_data, out_data, pre, n, post);
   } else {
-    lite::arm::math::elementwise_sub(
+    lite::arm::math::elementwise_sub<T>(
         x_data, y_data, out_data, x_dims.production());
   }
 }
@@ -419,12 +420,10 @@ REGISTER_LITE_KERNEL(
     .BindOutput("Out", {LiteType::GetTensorTy(TARGET(kARM))})
     .Finalize();
 
-REGISTER_LITE_KERNEL(elementwise_sub,
-                     kARM,
-                     kFloat,
-                     kNCHW,
-                     paddle::lite::kernels::arm::ElementwiseSubCompute,
-                     def)
+using elementwise_sub_float_t =
+    paddle::lite::kernels::arm::ElementwiseSubCompute<float, PRECISION(kFloat)>;
+REGISTER_LITE_KERNEL(
+    elementwise_sub, kARM, kFloat, kNCHW, elementwise_sub_float_t, def)
     .BindInput("X", {LiteType::GetTensorTy(TARGET(kARM))})
     .BindInput("Y", {LiteType::GetTensorTy(TARGET(kARM))})
     .BindOutput("Out", {LiteType::GetTensorTy(TARGET(kARM))})
diff --git a/lite/kernels/arm/elementwise_compute.h b/lite/kernels/arm/elementwise_compute.h
index 025ce7cc7f677d0c20b3ff72a0604409810d0deb..46efac084e52e52431604da372c9061f2b07acb4 100644
--- a/lite/kernels/arm/elementwise_compute.h
+++ b/lite/kernels/arm/elementwise_compute.h
@@ -38,8 +38,8 @@ class ElementwiseAddActivationCompute
   virtual ~ElementwiseAddActivationCompute() = default;
 };
 
-class ElementwiseSubCompute
-    : public KernelLite<TARGET(kARM), PRECISION(kFloat)> {
+template <typename T, PrecisionType PType>
+class ElementwiseSubCompute : public KernelLite<TARGET(kARM), PType> {
  public:
   void Run() override;
 
diff --git a/lite/tests/kernels/elementwise_grad_compute_test.cc b/lite/tests/kernels/elementwise_grad_compute_test.cc
index baf7b16a94d4ba539bcd17d9c1b670001956a889..46964485cd1c80f4f79f401d6d017af8b0b66ebe 100644
--- a/lite/tests/kernels/elementwise_grad_compute_test.cc
+++ b/lite/tests/kernels/elementwise_grad_compute_test.cc
@@ -27,7 +27,7 @@ using param_t = operators::ElementwiseParam;
 using grad_param_t = operators::ElementwiseGradParam;
 using kernel_add_t = ElementwiseAddCompute<float, PRECISION(kFloat)>;
 using grad_kernel_add_t = ElementwiseAddGradCompute;
-using kernel_sub_t = ElementwiseSubCompute;
+using kernel_sub_t = ElementwiseSubCompute<float, PRECISION(kFloat)>;
 using grad_kernel_sub_t = ElementwiseSubGradCompute;
 
 void elementwise_common(grad_param_t& param,           // NOLINT