diff --git a/paddle/fluid/operators/activation_op_npu.cc b/paddle/fluid/operators/activation_op_npu.cc
old mode 100644
new mode 100755
index 02ce817bcc8b2bf99a3e6bf11dea444b21b6529a..5cf70cc391d8ffffe8e30cdb1d36797a75f89331
--- a/paddle/fluid/operators/activation_op_npu.cc
+++ b/paddle/fluid/operators/activation_op_npu.cc
@@ -560,6 +560,37 @@ class AtanGradNPUKernel : public framework::OpKernel<T> {
   }
 };
 
+template <typename DeviceContext, typename T>
+class ExpNPUKernel : public framework::OpKernel<T> {
+ public:
+  void Compute(const framework::ExecutionContext& ctx) const override {
+    auto* x = ctx.Input<framework::Tensor>("X");
+    auto* out = ctx.Output<framework::Tensor>("Out");
+    out->mutable_data<T>(ctx.GetPlace());
+    const auto& runner = NpuOpRunner("Exp", {*x}, {*out}, {});
+    auto stream =
+        ctx.template device_context<paddle::platform::NPUDeviceContext>()
+            .stream();
+    runner.Run(stream);
+  }
+};
+
+template <typename DeviceContext, typename T>
+class ExpGradNPUKernel : public framework::OpKernel<T> {
+ public:
+  void Compute(const framework::ExecutionContext& ctx) const override {
+    auto* out = ctx.Input<Tensor>("Out");
+    auto* dout = ctx.Input<Tensor>(framework::GradVarName("Out"));
+    auto* dx = ctx.Output<Tensor>(framework::GradVarName("X"));
+    dx->mutable_data<T>(ctx.GetPlace());
+    auto stream =
+        ctx.template device_context<paddle::platform::NPUDeviceContext>()
+            .stream();
+    const auto& runner = NpuOpRunner("Mul", {*dout, *out}, {*dx}, {});
+    runner.Run(stream);
+  }
+};
+
 }  // namespace operators
 }  // namespace paddle
 
@@ -692,3 +723,11 @@ REGISTER_OP_NPU_KERNEL(
     ops::AtanGradNPUKernel<paddle::platform::NPUDeviceContext, float>,
     ops::AtanGradNPUKernel<paddle::platform::NPUDeviceContext,
                            paddle::platform::float16>);
+
+REGISTER_OP_NPU_KERNEL(
+    exp, ops::ExpNPUKernel<paddle::platform::NPUDeviceContext, float>,
+    ops::ExpNPUKernel<paddle::platform::NPUDeviceContext, double>);
+
+REGISTER_OP_NPU_KERNEL(
+    exp_grad, ops::ExpGradNPUKernel<paddle::platform::NPUDeviceContext, float>,
+    ops::ExpGradNPUKernel<paddle::platform::NPUDeviceContext, double>);
diff --git a/python/paddle/fluid/tests/unittests/npu/test_exp_op_npu.py b/python/paddle/fluid/tests/unittests/npu/test_exp_op_npu.py
new file mode 100755
index 0000000000000000000000000000000000000000..ccd5f0649d8dc68bb9cc8bb3e1736ced26c7cf7f
--- /dev/null
+++ b/python/paddle/fluid/tests/unittests/npu/test_exp_op_npu.py
@@ -0,0 +1,73 @@
+# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
+# 
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+# 
+#     http://www.apache.org/licenses/LICENSE-2.0
+# 
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from __future__ import print_function
+import unittest
+
+import numpy as np
+from scipy.special import expit, erf
+
+from paddle.fluid.tests.unittests.op_test import OpTest, convert_float_to_uint16
+import paddle
+import paddle.nn as nn
+import paddle.nn.functional as F
+import paddle.fluid as fluid
+import paddle.fluid.core as core
+from paddle.fluid import compiler, Program, program_guard
+
+paddle.enable_static()
+SEED = 2049
+
+
+class TestExpNPUOP(OpTest):
+    def setUp(self):
+
+        self.set_npu()
+        self.place = paddle.NPUPlace(0)
+        self.op_type = "exp"
+        self.init_dtype()
+        self.init_kernel_type()
+
+        np.random.seed(SEED)
+        x = np.random.uniform(0.1, 1, [11, 17]).astype(self.dtype)
+        out = np.exp(x)
+
+        self.inputs = {'X': OpTest.np_dtype_to_fluid_dtype(x)}
+        self.outputs = {'Out': out}
+
+    def test_check_output(self):
+        self.check_output_with_place(self.place)
+
+    def test_check_grad(self):
+        if self.dtype == np.float16:
+            return
+        self.check_grad_with_place(self.place, ['X'], 'Out')
+
+    def init_dtype(self):
+        self.dtype = np.float32
+
+    def init_kernel_type(self):
+        pass
+
+    def set_npu(self):
+        self.__class__.use_npu = True
+
+
+class TestExpNPUOPFloat64(TestExpNPUOP):
+    def init_dtype(self):
+        self.dtype = np.float64
+
+
+if __name__ == "__main__":
+    unittest.main()