diff --git a/.clang-tidy b/.clang-tidy
index 6371a8a89..37f90618f 100644
--- a/.clang-tidy
+++ b/.clang-tidy
@@ -10,28 +10,34 @@ cppcoreguidelines-*,\
 -cppcoreguidelines-avoid-c-arrays,\
 -cppcoreguidelines-avoid-non-const-global-variables,\
 -cppcoreguidelines-macro-usage,\
+-cppcoreguidelines-missing-std-forward,\
 -cppcoreguidelines-owning-memory,\
 -cppcoreguidelines-pro-bounds-pointer-arithmetic,\
 -cppcoreguidelines-pro-type-const-cast,\
 -cppcoreguidelines-pro-type-vararg,\
+-cppcoreguidelines-rvalue-reference-param-not-moved,\
 hicpp-*,\
 -hicpp-avoid-c-arrays,\
 -hicpp-vararg,\
 -hicpp-use-auto,\
 misc-*,\
 -misc-const-correctness,\
+-misc-include-cleaner,\
 modernize-*,\
 -modernize-avoid-c-arrays,\
 -modernize-use-auto,\
 -modernize-use-nodiscard,\
 -modernize-use-trailing-return-type,\
 performance-*,\
+-performance-avoid-endl,\
 portability-*,\
 readability-*,\
 -readability-avoid-const-params-in-decls,\
 -readability-const-return-type,\
 -readability-function-cognitive-complexity,\
 -readability-identifier-length,\
+-readability-redundant-casting,\
+-readability-redundant-member-init,\
 "
 HeaderFilterRegex: 'src|benchmark/stdgpu|test/stdgpu'
 ...
diff --git a/src/stdgpu/atomic.cuh b/src/stdgpu/atomic.cuh
index 76b00d5fa..b1bd86fa9 100644
--- a/src/stdgpu/atomic.cuh
+++ b/src/stdgpu/atomic.cuh
@@ -28,6 +28,7 @@
  */
 
 #include <cstddef>
+#include <cstdint>
 #include <type_traits>
 
 #include <stdgpu/execution.h>
@@ -42,16 +43,52 @@ namespace stdgpu
  * \ingroup atomic
  * \brief The memory order types for atomic operations
  */
-enum memory_order
+enum class memory_order : std::int8_t
 {
-    memory_order_relaxed, /**< memory_order_relaxed */
-    memory_order_consume, /**< memory_order_consume */
-    memory_order_acquire, /**< memory_order_acquire */
-    memory_order_release, /**< memory_order_release */
-    memory_order_acq_rel, /**< memory_order_acq_rel */
-    memory_order_seq_cst  /**< memory_order_seq_cst */
+    relaxed, /**< relaxed */
+    consume, /**< consume */
+    acquire, /**< acquire */
+    release, /**< release */
+    acq_rel, /**< acq_rel */
+    seq_cst  /**< seq_cst */
 };
 
+/**
+ * \ingroup atomic
+ * \brief memory_order_relaxed
+ */
+inline constexpr memory_order memory_order_relaxed = memory_order::relaxed;
+
+/**
+ * \ingroup atomic
+ * \brief memory_order_consume
+ */
+inline constexpr memory_order memory_order_consume = memory_order::consume;
+
+/**
+ * \ingroup atomic
+ * \brief memory_order_acquire
+ */
+inline constexpr memory_order memory_order_acquire = memory_order::acquire;
+
+/**
+ * \ingroup atomic
+ * \brief memory_order_release
+ */
+inline constexpr memory_order memory_order_release = memory_order::release;
+
+/**
+ * \ingroup atomic
+ * \brief memory_order_acq_rel
+ */
+inline constexpr memory_order memory_order_acq_rel = memory_order::acq_rel;
+
+/**
+ * \ingroup atomic
+ * \brief memory_order_seq_cst
+ */
+inline constexpr memory_order memory_order_seq_cst = memory_order::seq_cst;
+
 /**
  * \ingroup atomic
  * \brief A synchronization fence enforcing the given memory order
diff --git a/src/stdgpu/impl/algorithm_detail.h b/src/stdgpu/impl/algorithm_detail.h
index fd816d328..9f275fd81 100644
--- a/src/stdgpu/impl/algorithm_detail.h
+++ b/src/stdgpu/impl/algorithm_detail.h
@@ -45,7 +45,7 @@ clamp(const T& v, const T& lower, const T& upper)
 {
     STDGPU_EXPECTS(!(upper < lower));
 
-    return v < lower ? lower : upper < v ? upper : v;
+    return v < lower ? lower : upper < v ? upper : v; // NOLINT(readability-avoid-nested-conditional-operator)
 }
 
 template <typename IndexType,
diff --git a/src/stdgpu/impl/bitset_detail.cuh b/src/stdgpu/impl/bitset_detail.cuh
index e9944d1d0..c5c8fb73a 100644
--- a/src/stdgpu/impl/bitset_detail.cuh
+++ b/src/stdgpu/impl/bitset_detail.cuh
@@ -130,20 +130,20 @@ template <typename Block>
 class count_block_bits
 {
 public:
-    inline count_block_bits(Block* bit_blocks, const index_t size)
+    count_block_bits(Block* bit_blocks, const index_t size)
       : _bit_blocks(bit_blocks)
       , _size(size)
     {
     }
 
-    inline STDGPU_HOST_DEVICE index_t
+    STDGPU_HOST_DEVICE index_t
     operator()(const index_t i) const
     {
         return static_cast<index_t>(popcount(block_mask(i) & _bit_blocks[i]));
     }
 
 private:
-    inline STDGPU_HOST_DEVICE Block
+    STDGPU_HOST_DEVICE Block
     block_mask(const index_t i) const
     {
         index_t remaining_bits = _size - i * _bits_per_block;
@@ -162,12 +162,12 @@ template <typename Block>
 class flip_bits
 {
 public:
-    inline explicit flip_bits(Block* bit_blocks)
+    explicit flip_bits(Block* bit_blocks)
       : _bit_blocks(bit_blocks)
     {
     }
 
-    inline STDGPU_HOST_DEVICE void
+    STDGPU_HOST_DEVICE void
     operator()(const index_t i)
     {
         _bit_blocks[i] = ~_bit_blocks[i];
diff --git a/src/stdgpu/impl/functional_detail.h b/src/stdgpu/impl/functional_detail.h
index 27b4a9aee..fcb733a2c 100644
--- a/src/stdgpu/impl/functional_detail.h
+++ b/src/stdgpu/impl/functional_detail.h
@@ -48,7 +48,7 @@ struct hash_base
 template <typename T>
 struct hash_base<T, true>
 {
-    inline STDGPU_HOST_DEVICE std::size_t
+    STDGPU_HOST_DEVICE std::size_t
     operator()(const T& key) const
     {
         return hash<std::underlying_type_t<T>>()(static_cast<std::underlying_type_t<T>>(key));
diff --git a/src/stdgpu/impl/iterator_detail.h b/src/stdgpu/impl/iterator_detail.h
index ba3442f4d..513d2e5fd 100644
--- a/src/stdgpu/impl/iterator_detail.h
+++ b/src/stdgpu/impl/iterator_detail.h
@@ -42,6 +42,7 @@ template <typename T>
 index64_t
 size(T* array)
 {
+    // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
     index64_t array_size_bytes = size<void>(static_cast<void*>(const_cast<std::remove_cv_t<T>*>(array)));
 
     if (array_size_bytes % static_cast<index64_t>(sizeof(T)) != 0) // NOLINT(bugprone-sizeof-expression)
diff --git a/src/stdgpu/impl/memory.cpp b/src/stdgpu/impl/memory.cpp
index 68b1b9596..dd027d3f6 100644
--- a/src/stdgpu/impl/memory.cpp
+++ b/src/stdgpu/impl/memory.cpp
@@ -15,6 +15,7 @@
 
 #include <stdgpu/memory.h>
 
+#include <cstdint>
 #include <cstdio>
 #include <map>
 #include <mutex>
diff --git a/src/stdgpu/impl/memory_detail.h b/src/stdgpu/impl/memory_detail.h
index e2cc86911..8b54bf1bb 100644
--- a/src/stdgpu/impl/memory_detail.h
+++ b/src/stdgpu/impl/memory_detail.h
@@ -377,12 +377,15 @@ copyDevice2HostArray(const T* source_device_array,
                      T* destination_host_array,
                      const MemoryCopy check_safety)
 {
-    stdgpu::detail::memcpy(destination_host_array,
-                           source_device_array,
-                           count * static_cast<stdgpu::index64_t>(sizeof(T)), // NOLINT(bugprone-sizeof-expression)
-                           stdgpu::dynamic_memory_type::host,
-                           stdgpu::dynamic_memory_type::device,
-                           check_safety != MemoryCopy::RANGE_CHECK);
+    stdgpu::detail::memcpy(
+            // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
+            static_cast<void*>(const_cast<std::remove_cv_t<T>*>(destination_host_array)),
+            // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
+            static_cast<void*>(const_cast<std::remove_cv_t<T>*>(source_device_array)),
+            count * static_cast<stdgpu::index64_t>(sizeof(T)), // NOLINT(bugprone-sizeof-expression)
+            stdgpu::dynamic_memory_type::host,
+            stdgpu::dynamic_memory_type::device,
+            check_safety != MemoryCopy::RANGE_CHECK);
 }
 
 template <typename T>
@@ -392,12 +395,15 @@ copyHost2DeviceArray(const T* source_host_array,
                      T* destination_device_array,
                      const MemoryCopy check_safety)
 {
-    stdgpu::detail::memcpy(destination_device_array,
-                           source_host_array,
-                           count * static_cast<stdgpu::index64_t>(sizeof(T)), // NOLINT(bugprone-sizeof-expression)
-                           stdgpu::dynamic_memory_type::device,
-                           stdgpu::dynamic_memory_type::host,
-                           check_safety != MemoryCopy::RANGE_CHECK);
+    stdgpu::detail::memcpy(
+            // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
+            static_cast<void*>(const_cast<std::remove_cv_t<T>*>(destination_device_array)),
+            // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
+            static_cast<void*>(const_cast<std::remove_cv_t<T>*>(source_host_array)),
+            count * static_cast<stdgpu::index64_t>(sizeof(T)), // NOLINT(bugprone-sizeof-expression)
+            stdgpu::dynamic_memory_type::device,
+            stdgpu::dynamic_memory_type::host,
+            check_safety != MemoryCopy::RANGE_CHECK);
 }
 
 template <typename T>
@@ -407,12 +413,15 @@ copyHost2HostArray(const T* source_host_array,
                    T* destination_host_array,
                    const MemoryCopy check_safety)
 {
-    stdgpu::detail::memcpy(destination_host_array,
-                           source_host_array,
-                           count * static_cast<stdgpu::index64_t>(sizeof(T)), // NOLINT(bugprone-sizeof-expression)
-                           stdgpu::dynamic_memory_type::host,
-                           stdgpu::dynamic_memory_type::host,
-                           check_safety != MemoryCopy::RANGE_CHECK);
+    stdgpu::detail::memcpy(
+            // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
+            static_cast<void*>(const_cast<std::remove_cv_t<T>*>(destination_host_array)),
+            // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
+            static_cast<void*>(const_cast<std::remove_cv_t<T>*>(source_host_array)),
+            count * static_cast<stdgpu::index64_t>(sizeof(T)), // NOLINT(bugprone-sizeof-expression)
+            stdgpu::dynamic_memory_type::host,
+            stdgpu::dynamic_memory_type::host,
+            check_safety != MemoryCopy::RANGE_CHECK);
 }
 
 template <typename T>
@@ -422,12 +431,15 @@ copyDevice2DeviceArray(const T* source_device_array,
                        T* destination_device_array,
                        const MemoryCopy check_safety)
 {
-    stdgpu::detail::memcpy(destination_device_array,
-                           source_device_array,
-                           count * static_cast<stdgpu::index64_t>(sizeof(T)), // NOLINT(bugprone-sizeof-expression)
-                           stdgpu::dynamic_memory_type::device,
-                           stdgpu::dynamic_memory_type::device,
-                           check_safety != MemoryCopy::RANGE_CHECK);
+    stdgpu::detail::memcpy(
+            // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
+            static_cast<void*>(const_cast<std::remove_cv_t<T>*>(destination_device_array)),
+            // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
+            static_cast<void*>(const_cast<std::remove_cv_t<T>*>(source_device_array)),
+            count * static_cast<stdgpu::index64_t>(sizeof(T)), // NOLINT(bugprone-sizeof-expression)
+            stdgpu::dynamic_memory_type::device,
+            stdgpu::dynamic_memory_type::device,
+            check_safety != MemoryCopy::RANGE_CHECK);
 }
 
 namespace stdgpu
@@ -517,7 +529,8 @@ void
 safe_device_allocator<T>::deallocate(T* p, index64_t n)
 {
     deregister_memory(p, n, memory_type);
-    detail::deallocate(static_cast<void*>(p),
+    // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
+    detail::deallocate(static_cast<void*>(const_cast<std::remove_cv_t<T>*>(p)),
                        n * static_cast<index64_t>(sizeof(T)), // NOLINT(bugprone-sizeof-expression)
                        memory_type);
 }
@@ -543,7 +556,8 @@ void
 safe_host_allocator<T>::deallocate(T* p, index64_t n)
 {
     deregister_memory(p, n, memory_type);
-    detail::deallocate(static_cast<void*>(p),
+    // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
+    detail::deallocate(static_cast<void*>(const_cast<std::remove_cv_t<T>*>(p)),
                        n * static_cast<index64_t>(sizeof(T)), // NOLINT(bugprone-sizeof-expression)
                        memory_type);
 }
@@ -569,7 +583,8 @@ void
 safe_managed_allocator<T>::deallocate(T* p, index64_t n)
 {
     deregister_memory(p, n, memory_type);
-    detail::deallocate(static_cast<void*>(p),
+    // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
+    detail::deallocate(static_cast<void*>(const_cast<std::remove_cv_t<T>*>(p)),
                        n * static_cast<index64_t>(sizeof(T)), // NOLINT(bugprone-sizeof-expression)
                        memory_type);
 }
@@ -695,7 +710,8 @@ template <typename T, typename... Args>
 STDGPU_HOST_DEVICE T*
 construct_at(T* p, Args&&... args)
 {
-    return ::new (static_cast<void*>(p)) T(forward<Args>(args)...);
+    // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
+    return ::new (static_cast<void*>(const_cast<std::remove_cv_t<T>*>(p))) T(forward<Args>(args)...);
 }
 
 template <typename T>
@@ -793,6 +809,7 @@ template <typename T>
 dynamic_memory_type
 get_dynamic_memory_type(T* array)
 {
+    // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
     return get_dynamic_memory_type<void>(static_cast<void*>(const_cast<std::remove_cv_t<T>*>(array)));
 }
 
@@ -804,6 +821,7 @@ template <typename T>
 void
 register_memory(T* p, index64_t n, dynamic_memory_type memory_type)
 {
+    // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
     register_memory<void>(static_cast<void*>(const_cast<std::remove_cv_t<T>*>(p)),
                           n * static_cast<index64_t>(sizeof(T)), // NOLINT(bugprone-sizeof-expression)
                           memory_type);
@@ -817,6 +835,7 @@ template <typename T>
 void
 deregister_memory(T* p, index64_t n, dynamic_memory_type memory_type)
 {
+    // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
     deregister_memory<void>(static_cast<void*>(const_cast<std::remove_cv_t<T>*>(p)),
                             n * static_cast<index64_t>(sizeof(T)), // NOLINT(bugprone-sizeof-expression)
                             memory_type);
@@ -830,6 +849,7 @@ template <typename T>
 index64_t
 size_bytes(T* array)
 {
+    // NOLINTNEXTLINE(bugprone-multi-level-implicit-pointer-conversion)
     return size_bytes<void>(static_cast<void*>(const_cast<std::remove_cv_t<T>*>(array)));
 }
 
diff --git a/src/stdgpu/impl/unordered_base.cuh b/src/stdgpu/impl/unordered_base.cuh
index f9605a315..ca6b538e4 100644
--- a/src/stdgpu/impl/unordered_base.cuh
+++ b/src/stdgpu/impl/unordered_base.cuh
@@ -16,6 +16,8 @@
 #ifndef STDGPU_UNORDERED_BASE_H
 #define STDGPU_UNORDERED_BASE_H
 
+#include <cstdint>
+
 #include <stdgpu/atomic.cuh>
 #include <stdgpu/bitset.cuh>
 #include <stdgpu/cstddef.h>
@@ -35,7 +37,7 @@ namespace stdgpu::detail
 /**
  * \brief Status flags for try_insert and try_erase
  */
-enum class operation_status
+enum class operation_status : std::int8_t
 {
     success,                   /**< Operation succeeded */
     failed_no_action_required, /**< Operation failed because no action is required */
diff --git a/src/stdgpu/memory.h b/src/stdgpu/memory.h
index 7a5f0d81b..489cc6542 100644
--- a/src/stdgpu/memory.h
+++ b/src/stdgpu/memory.h
@@ -25,6 +25,7 @@
  * \file stdgpu/memory.h
  */
 
+#include <cstdint>
 #include <functional>
 #include <memory>
 #include <type_traits>
@@ -39,7 +40,7 @@
  * \ingroup memory
  * \brief The place to initialize the created array
  */
-enum class Initialization
+enum class Initialization : std::int8_t
 {
     HOST,  /**< The array is initialized on the host (CPU) */
     DEVICE /**< The array is initialized on the device (GPU) */
@@ -121,7 +122,7 @@ destroyManagedArray(T*& managed_array);
  * \ingroup memory
  * \brief The copy check states
  */
-enum class MemoryCopy
+enum class MemoryCopy : std::int8_t
 {
     NO_CHECK,   /**< No checks should be performed. This is useful when copying from/to arrays not created by our API,
                    e.g. created by 3rd party libraries or pointers to local variables. */
@@ -356,7 +357,7 @@ class device_unique_object
  * \ingroup memory
  * \brief The types of a dynamically allocated array
  */
-enum class dynamic_memory_type
+enum class dynamic_memory_type : std::int8_t
 {
     host,    /**< The array is allocated on the host (CPU) */
     device,  /**< The array is allocated on the device (GPU) */
diff --git a/src/stdgpu/utility.h b/src/stdgpu/utility.h
index 04c2afae5..a81f59cf4 100644
--- a/src/stdgpu/utility.h
+++ b/src/stdgpu/utility.h
@@ -182,9 +182,9 @@ struct pair
     constexpr STDGPU_HOST_DEVICE pair&
     operator=(pair<U1, U2>&& p);
 
-    // NOLINTNEXTLINE(misc-non-private-member-variables-in-classes)
+    // NOLINTNEXTLINE(cppcoreguidelines-avoid-const-or-ref-data-members,misc-non-private-member-variables-in-classes)
     first_type first; /**< First element of pair */
-    // NOLINTNEXTLINE(misc-non-private-member-variables-in-classes)
+    // NOLINTNEXTLINE(cppcoreguidelines-avoid-const-or-ref-data-members,misc-non-private-member-variables-in-classes)
     second_type second; /**< Second element of pair */
 };
 
diff --git a/tests/stdgpu/functional.cpp b/tests/stdgpu/functional.cpp
index 4860ed928..2a675fb7d 100644
--- a/tests/stdgpu/functional.cpp
+++ b/tests/stdgpu/functional.cpp
@@ -15,6 +15,7 @@
 
 #include <gtest/gtest.h>
 
+#include <cstdint>
 #include <limits>
 #include <random>
 #include <unordered_set>
@@ -247,7 +248,7 @@ TEST_F(stdgpu_functional, hash_long_double)
     hash_check_floating_point_random<long double>();
 }
 
-enum old_enum
+enum old_enum : std::int8_t
 {
     zero = 0,
     one = 1,
@@ -270,7 +271,7 @@ TEST_F(stdgpu_functional, hash_enum)
     EXPECT_GT(static_cast<stdgpu::index_t>(hashes.size()), 4 * 90 / 100);
 }
 
-enum class scoped_enum
+enum class scoped_enum : std::int8_t
 {
     zero = 0,
     one = 1,
diff --git a/tests/stdgpu/iterator.cpp b/tests/stdgpu/iterator.cpp
index a4c8e3d23..05a07daed 100644
--- a/tests/stdgpu/iterator.cpp
+++ b/tests/stdgpu/iterator.cpp
@@ -347,7 +347,7 @@ class back_insert_interface
     }
 
 private:
-    std::vector<int>& _vector;
+    std::vector<int>& _vector; // NOLINT(cppcoreguidelines-avoid-const-or-ref-data-members)
 };
 
 class front_insert_interface
@@ -367,7 +367,7 @@ class front_insert_interface
     }
 
 private:
-    std::vector<int>& _vector;
+    std::vector<int>& _vector; // NOLINT(cppcoreguidelines-avoid-const-or-ref-data-members)
 };
 
 class insert_interface
@@ -387,7 +387,7 @@ class insert_interface
     }
 
 private:
-    std::vector<int>& _vector;
+    std::vector<int>& _vector; // NOLINT(cppcoreguidelines-avoid-const-or-ref-data-members)
 };
 
 TEST_F(stdgpu_iterator, back_inserter)
diff --git a/tests/stdgpu/limits.cpp b/tests/stdgpu/limits.cpp
index 1bdaaed8c..1e9d9e1f5 100644
--- a/tests/stdgpu/limits.cpp
+++ b/tests/stdgpu/limits.cpp
@@ -210,7 +210,7 @@ TEST_F(stdgpu_limits, long_double)
 class NonArithmeticType
 {
 public:
-    inline STDGPU_HOST_DEVICE bool
+    STDGPU_HOST_DEVICE bool
     operator==(const NonArithmeticType& other) const
     {
         return x == other.x;