diff --git a/tests/validation_new/CPP/L2Normalize.cpp b/tests/validation_new/CPP/L2Normalize.cpp
new file mode 100644
index 0000000..f423702
--- /dev/null
+++ b/tests/validation_new/CPP/L2Normalize.cpp
@@ -0,0 +1,88 @@
+/*
+ * Copyright (c) 2017 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include "L2Normalize.h"
+#include "ReductionOperation.h"
+
+#include "tests/validation_new/Helpers.h"
+
+#include <algorithm>
+#include <cmath>
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+namespace reference
+{
+namespace
+{
+TensorShape get_output_shape(TensorShape shape, unsigned int axis)
+{
+    TensorShape output_shape(shape);
+    output_shape.set(axis, 1);
+    return output_shape;
+}
+} // namespace
+
+template <typename T>
+SimpleTensor<T> l2_normalize(const SimpleTensor<T> &src, unsigned int axis, float epsilon)
+{
+    // Create reference
+    SimpleTensor<T> dst{ src.shape(), src.data_type() };
+
+    // Reduce across given axis
+    SimpleTensor<T> sum = reduction_operation(src, get_output_shape(src.shape(), axis), axis, ReductionOperation::SUM_SQUARE);
+
+    // Compute reference
+    const int elems      = src.shape()[axis];
+    const int upper_dims = src.shape().total_size_upper(axis + 1);
+
+    for(int du = 0; du < upper_dims; ++du)
+    {
+        if(axis == 0)
+        {
+            const T *src_row_ptr         = src.data() + du * elems;
+            T       *dst_row_ptr         = dst.data() + du * elems;
+            const T  normalization_value = std::sqrt(std::max(sum[du], epsilon));
+            std::transform(src_row_ptr, src_row_ptr + elems, dst_row_ptr, [normalization_value](T val)
+            {
+                return val / normalization_value;
+            });
+        }
+        else
+        {
+            ARM_COMPUTE_ERROR("Unsupported normalization axis");
+        }
+    }
+
+    return dst;
+}
+
+template SimpleTensor<float> l2_normalize(const SimpleTensor<float> &src, unsigned int axis, float epsilon);
+} // namespace reference
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
diff --git a/tests/validation_new/CPP/L2Normalize.h b/tests/validation_new/CPP/L2Normalize.h
new file mode 100644
index 0000000..3e81b9a
--- /dev/null
+++ b/tests/validation_new/CPP/L2Normalize.h
@@ -0,0 +1,44 @@
+/*
+ * Copyright (c) 2017 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef __ARM_COMPUTE_TEST_L2NORMALIZE_H__
+#define __ARM_COMPUTE_TEST_L2NORMALIZE_H__
+
+#include "tests/validation_new/Helpers.h"
+#include "tests/validation_new/SimpleTensor.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+namespace reference
+{
+template <typename T>
+SimpleTensor<T> l2_normalize(const SimpleTensor<T> &src, unsigned int axis, float epsilon);
+} // namespace reference
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
+#endif /* __ARM_COMPUTE_TEST_L2NORMALIZE_H__ */
diff --git a/tests/validation_new/CPP/ReductionOperation.cpp b/tests/validation_new/CPP/ReductionOperation.cpp
new file mode 100644
index 0000000..f4c5c21
--- /dev/null
+++ b/tests/validation_new/CPP/ReductionOperation.cpp
@@ -0,0 +1,93 @@
+/*
+ * Copyright (c) 2017 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include "ReductionOperation.h"
+
+#include "tests/validation_new/Helpers.h"
+
+#include <algorithm>
+#include <cmath>
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+namespace reference
+{
+namespace
+{
+template <typename T>
+struct square
+{
+    T operator()(const T &lhs, const T &rhs) const
+    {
+        return (lhs + rhs * rhs);
+    }
+};
+
+template <typename T>
+T reduce_operation(T *ptr, int reduce_elements, ReductionOperation op)
+{
+    switch(op)
+    {
+        case ReductionOperation::SUM_SQUARE:
+            return std::accumulate(ptr, ptr + reduce_elements, 0.f, square<T>());
+        default:
+            ARM_COMPUTE_ERROR("Unsupported reduction operation");
+    }
+}
+} // namespace
+
+template <typename T>
+SimpleTensor<T> reduction_operation(const SimpleTensor<T> &src, const TensorShape &dst_shape, unsigned int axis, ReductionOperation op)
+{
+    // Create reference
+    SimpleTensor<T> dst{ dst_shape, src.data_type() };
+
+    // Compute reference
+    const int reduce_elems = src.shape()[axis];
+    const int upper_dims   = src.shape().total_size_upper(axis + 1);
+
+    for(int du = 0; du < upper_dims; ++du)
+    {
+        if(axis == 0)
+        {
+            const T *src_row_ptr = src.data() + du * reduce_elems;
+            dst[du]              = reduce_operation(src_row_ptr, reduce_elems, op);
+        }
+        else
+        {
+            ARM_COMPUTE_ERROR("Unsupported reduction axis");
+        }
+    }
+
+    return dst;
+}
+
+template SimpleTensor<float> reduction_operation(const SimpleTensor<float> &src, const TensorShape &dst_shape, unsigned int axis, ReductionOperation op);
+} // namespace reference
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
diff --git a/tests/validation_new/CPP/ReductionOperation.h b/tests/validation_new/CPP/ReductionOperation.h
new file mode 100644
index 0000000..43a700d
--- /dev/null
+++ b/tests/validation_new/CPP/ReductionOperation.h
@@ -0,0 +1,44 @@
+/*
+ * Copyright (c) 2017 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef __ARM_COMPUTE_TEST_REDUCTION_OPERATION_H__
+#define __ARM_COMPUTE_TEST_REDUCTION_OPERATION_H__
+
+#include "tests/validation_new/Helpers.h"
+#include "tests/validation_new/SimpleTensor.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+namespace reference
+{
+template <typename T>
+SimpleTensor<T> reduction_operation(const SimpleTensor<T> &src, const TensorShape &dst_shape, unsigned int axis, ReductionOperation op);
+} // namespace reference
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
+#endif /* __ARM_COMPUTE_TEST_FLOOR_H__ */
diff --git a/tests/validation_new/NEON/L2Normalize.cpp b/tests/validation_new/NEON/L2Normalize.cpp
new file mode 100644
index 0000000..20804d2
--- /dev/null
+++ b/tests/validation_new/NEON/L2Normalize.cpp
@@ -0,0 +1,75 @@
+/*
+ * Copyright (c) 2017 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include "arm_compute/core/Types.h"
+#include "arm_compute/runtime/NEON/functions/NEL2Normalize.h"
+#include "arm_compute/runtime/Tensor.h"
+#include "arm_compute/runtime/TensorAllocator.h"
+#include "framework/Asserts.h"
+#include "framework/Macros.h"
+#include "framework/datasets/Datasets.h"
+#include "tests/NEON/Accessor.h"
+#include "tests/PaddingCalculator.h"
+#include "tests/datasets_new/ShapeDatasets.h"
+#include "tests/validation_new/Validation.h"
+#include "tests/validation_new/fixtures/L2NormalizeFixture.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+namespace
+{
+/** Tolerance for float operations */
+constexpr RelativeTolerance tolerance_f32(0.00001f);
+} // namespace
+
+TEST_SUITE(NEON)
+TEST_SUITE(L2Normalize)
+
+template <typename T>
+using NEL2NormalizeFixture = L2NormalizeValidationFixture<Tensor, Accessor, NEL2Normalize, T>;
+
+TEST_SUITE(FP32)
+FIXTURE_DATA_TEST_CASE(RunSmall, NEL2NormalizeFixture<float>, framework::DatasetMode::PRECOMMIT,
+                       combine(combine(combine(datasets::SmallShapes(), framework::dataset::make("DataType", DataType::F32)), framework::dataset::make("Axis", { 0 })), framework::dataset::make("Epsilon", { 1e-12 })))
+{
+    // Validate output
+    validate(Accessor(_target), _reference, tolerance_f32);
+}
+
+FIXTURE_DATA_TEST_CASE(RunLarge, NEL2NormalizeFixture<float>, framework::DatasetMode::NIGHTLY,
+                       combine(combine(combine(datasets::LargeShapes(), framework::dataset::make("DataType", DataType::F32)), framework::dataset::make("Axis", { 0 })), framework::dataset::make("Epsilon", { 1e-12 })))
+{
+    // Validate output
+    validate(Accessor(_target), _reference, tolerance_f32);
+}
+TEST_SUITE_END()
+
+TEST_SUITE_END()
+TEST_SUITE_END()
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
diff --git a/tests/validation_new/NEON/ReductionOperation.cpp b/tests/validation_new/NEON/ReductionOperation.cpp
new file mode 100644
index 0000000..4048af0
--- /dev/null
+++ b/tests/validation_new/NEON/ReductionOperation.cpp
@@ -0,0 +1,75 @@
+/*
+ * Copyright (c) 2017 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include "arm_compute/core/Types.h"
+#include "arm_compute/runtime/NEON/functions/NEReductionOperation.h"
+#include "arm_compute/runtime/Tensor.h"
+#include "arm_compute/runtime/TensorAllocator.h"
+#include "framework/Asserts.h"
+#include "framework/Macros.h"
+#include "framework/datasets/Datasets.h"
+#include "tests/NEON/Accessor.h"
+#include "tests/PaddingCalculator.h"
+#include "tests/datasets_new/ReductionOperationDataset.h"
+#include "tests/datasets_new/ShapeDatasets.h"
+#include "tests/validation_new/Validation.h"
+#include "tests/validation_new/fixtures/ReductionOperationFixture.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+namespace
+{
+/** Tolerance for float operations */
+constexpr RelativeTolerance tolerance_f32(0.00001f);
+} // namespace
+
+TEST_SUITE(NEON)
+TEST_SUITE(ReductionOperation)
+
+template <typename T>
+using NEReductionOperationFixture = ReductionOperationValidationFixture<Tensor, Accessor, NEReductionOperation, T>;
+
+TEST_SUITE(FP32)
+FIXTURE_DATA_TEST_CASE(RunSmall, NEReductionOperationFixture<float>, framework::DatasetMode::PRECOMMIT,
+                       combine(combine(combine(datasets::SmallShapes(), framework::dataset::make("DataType", DataType::F32)), framework::dataset::make("Axis", { 0 })), datasets::ReductionOperations()))
+{
+    // Validate output
+    validate(Accessor(_target), _reference, tolerance_f32);
+}
+FIXTURE_DATA_TEST_CASE(RunLarge, NEReductionOperationFixture<float>, framework::DatasetMode::NIGHTLY,
+                       combine(combine(combine(datasets::LargeShapes(), framework::dataset::make("DataType", DataType::F32)), framework::dataset::make("Axis", { 0 })), datasets::ReductionOperations()))
+{
+    // Validate output
+    validate(Accessor(_target), _reference, tolerance_f32);
+}
+TEST_SUITE_END()
+
+TEST_SUITE_END()
+TEST_SUITE_END()
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
diff --git a/tests/validation_new/fixtures/L2NormalizeFixture.h b/tests/validation_new/fixtures/L2NormalizeFixture.h
new file mode 100644
index 0000000..5102b93
--- /dev/null
+++ b/tests/validation_new/fixtures/L2NormalizeFixture.h
@@ -0,0 +1,107 @@
+/*
+ * Copyright (c) 2017 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef ARM_COMPUTE_TEST_L2NORMALIZE_FIXTURE
+#define ARM_COMPUTE_TEST_L2NORMALIZE_FIXTURE
+
+#include "arm_compute/core/TensorShape.h"
+#include "arm_compute/core/Types.h"
+#include "arm_compute/runtime/Tensor.h"
+#include "framework/Asserts.h"
+#include "framework/Fixture.h"
+#include "tests/AssetsLibrary.h"
+#include "tests/Globals.h"
+#include "tests/IAccessor.h"
+#include "tests/validation_new/CPP/L2Normalize.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+template <typename TensorType, typename AccessorType, typename FunctionType, typename T>
+class L2NormalizeValidationFixture : public framework::Fixture
+{
+public:
+    template <typename...>
+    void setup(TensorShape shape, DataType data_type, unsigned int axis, float epsilon)
+    {
+        _target    = compute_target(shape, data_type, axis, epsilon);
+        _reference = compute_reference(shape, data_type, axis, epsilon);
+    }
+
+protected:
+    template <typename U>
+    void fill(U &&tensor)
+    {
+        library->fill_tensor_uniform(tensor, 0);
+    }
+
+    TensorType compute_target(const TensorShape &shape, DataType data_type, unsigned int axis, float epsilon)
+    {
+        // Create tensors
+        TensorType src = create_tensor<TensorType>(shape, data_type);
+        TensorType dst = create_tensor<TensorType>(shape, data_type);
+
+        // Create and configure function
+        FunctionType l2_norm_func;
+        l2_norm_func.configure(&src, &dst, axis, epsilon);
+
+        ARM_COMPUTE_EXPECT(src.info()->is_resizable(), framework::LogLevel::ERRORS);
+        ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
+
+        // Allocate tensors
+        src.allocator()->allocate();
+        dst.allocator()->allocate();
+
+        ARM_COMPUTE_EXPECT(!src.info()->is_resizable(), framework::LogLevel::ERRORS);
+        ARM_COMPUTE_EXPECT(!dst.info()->is_resizable(), framework::LogLevel::ERRORS);
+
+        // Fill tensors
+        fill(AccessorType(src));
+
+        // Compute function
+        l2_norm_func.run();
+
+        return dst;
+    }
+
+    SimpleTensor<T> compute_reference(const TensorShape &shape, DataType data_type, unsigned int axis, float epsilon)
+    {
+        // Create reference
+        SimpleTensor<T> src{ shape, data_type };
+
+        // Fill reference
+        fill(src);
+
+        return reference::l2_normalize<T>(src, axis, epsilon);
+    }
+
+    TensorType      _target{};
+    SimpleTensor<T> _reference{};
+};
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
+#endif /* ARM_COMPUTE_TEST_L2NORMALIZE_FIXTURE */
diff --git a/tests/validation_new/fixtures/ReductionOperationFixture.h b/tests/validation_new/fixtures/ReductionOperationFixture.h
new file mode 100644
index 0000000..416aadf
--- /dev/null
+++ b/tests/validation_new/fixtures/ReductionOperationFixture.h
@@ -0,0 +1,116 @@
+/*
+ * Copyright (c) 2017 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef ARM_COMPUTE_TEST_REDUCTION_OPERATION_FIXTURE
+#define ARM_COMPUTE_TEST_REDUCTION_OPERATION_FIXTURE
+
+#include "arm_compute/core/TensorShape.h"
+#include "arm_compute/core/Types.h"
+#include "arm_compute/runtime/Tensor.h"
+#include "framework/Asserts.h"
+#include "framework/Fixture.h"
+#include "tests/AssetsLibrary.h"
+#include "tests/Globals.h"
+#include "tests/IAccessor.h"
+#include "tests/validation_new/CPP/ReductionOperation.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+template <typename TensorType, typename AccessorType, typename FunctionType, typename T>
+class ReductionOperationValidationFixture : public framework::Fixture
+{
+public:
+    template <typename...>
+    void setup(TensorShape shape, DataType data_type, unsigned int axis, ReductionOperation op)
+    {
+        const TensorShape output_shape = get_output_shape(shape, axis);
+        _target                        = compute_target(shape, output_shape, data_type, axis, op);
+        _reference                     = compute_reference(shape, output_shape, data_type, axis, op);
+    }
+
+protected:
+    template <typename U>
+    void fill(U &&tensor)
+    {
+        library->fill_tensor_uniform(tensor, 0);
+    }
+
+    TensorType compute_target(const TensorShape &src_shape, const TensorShape &dst_shape, DataType data_type, unsigned int axis, ReductionOperation op)
+    {
+        // Create tensors
+        TensorType src = create_tensor<TensorType>(src_shape, data_type);
+        TensorType dst = create_tensor<TensorType>(dst_shape, data_type);
+
+        // Create and configure function
+        FunctionType reduction_func;
+        reduction_func.configure(&src, &dst, axis, op);
+
+        ARM_COMPUTE_EXPECT(src.info()->is_resizable(), framework::LogLevel::ERRORS);
+        ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
+
+        // Allocate tensors
+        src.allocator()->allocate();
+        dst.allocator()->allocate();
+
+        ARM_COMPUTE_EXPECT(!src.info()->is_resizable(), framework::LogLevel::ERRORS);
+        ARM_COMPUTE_EXPECT(!dst.info()->is_resizable(), framework::LogLevel::ERRORS);
+
+        // Fill tensors
+        fill(AccessorType(src));
+
+        // Compute function
+        reduction_func.run();
+
+        return dst;
+    }
+
+    SimpleTensor<T> compute_reference(const TensorShape &src_shape, const TensorShape &dst_shape, DataType data_type, unsigned int axis, ReductionOperation op)
+    {
+        // Create reference
+        SimpleTensor<T> src{ src_shape, data_type };
+
+        // Fill reference
+        fill(src);
+
+        return reference::reduction_operation<T>(src, dst_shape, axis, op);
+    }
+
+    TensorType      _target{};
+    SimpleTensor<T> _reference{};
+
+private:
+    TensorShape get_output_shape(TensorShape shape, unsigned int axis)
+    {
+        TensorShape output_shape(shape);
+        output_shape.set(axis, 1);
+        return output_shape;
+    }
+};
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
+#endif /* ARM_COMPUTE_TEST_REDUCTION_OPERATION_FIXTURE */