cuda-cccl 0.1.3.1.0.dev1486__cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl

cuda/cccl/headers/include/cuda/std/__limits/numeric_limits_ext.h

@@ -0,0 +1,781 @@
+// -*- C++ -*-
+//===----------------------------------------------------------------------===//
+//
+// Part of libcu++, the C++ Standard Library for your entire system,
+// under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+// SPDX-FileCopyrightText: Copyright (c) 2025 NVIDIA CORPORATION & AFFILIATES.
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef _LIBCUDACXX___LIMITS_NUMERIC_LIMITS_EXT_H
+#define _LIBCUDACXX___LIMITS_NUMERIC_LIMITS_EXT_H
+
+#include <cuda/std/detail/__config>
+
+#if defined(_CCCL_IMPLICIT_SYSTEM_HEADER_GCC)
+#  pragma GCC system_header
+#elif defined(_CCCL_IMPLICIT_SYSTEM_HEADER_CLANG)
+#  pragma clang system_header
+#elif defined(_CCCL_IMPLICIT_SYSTEM_HEADER_MSVC)
+#  pragma system_header
+#endif // no system header
+
+#include <cuda/std/__bit/bit_cast.h>
+#include <cuda/std/__floating_point/storage.h>
+#include <cuda/std/__limits/numeric_limits.h>
+
+#include <cuda/std/__cccl/prologue.h>
+
+_LIBCUDACXX_BEGIN_NAMESPACE_STD
+
+// nvfp16
+
+#if _CCCL_HAS_NVFP16()
+template <>
+class __numeric_limits_impl<__half, __numeric_limits_type::__floating_point>
+{
+public:
+  using type = __half;
+
+  static constexpr bool is_specialized = true;
+
+  static constexpr bool is_signed   = true;
+  static constexpr int digits       = 11;
+  static constexpr int digits10     = 3;
+  static constexpr int max_digits10 = 5;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__half>(uint16_t(0x0400u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type max() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__half>(uint16_t(0x7bffu));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type lowest() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__half>(uint16_t(0xfbffu));
+  }
+
+  static constexpr bool is_integer = false;
+  static constexpr bool is_exact   = false;
+  static constexpr int radix       = FLT_RADIX;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type epsilon() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__half>(uint16_t(0x1400u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type round_error() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__half>(uint16_t(0x3800u));
+  }
+
+  static constexpr int min_exponent   = -13;
+  static constexpr int min_exponent10 = -4;
+  static constexpr int max_exponent   = 16;
+  static constexpr int max_exponent10 = 4;
+
+  static constexpr bool has_infinity                                             = true;
+  static constexpr bool has_quiet_NaN                                            = true;
+  static constexpr bool has_signaling_NaN                                        = true;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr float_denorm_style has_denorm = denorm_present;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr bool has_denorm_loss          = false;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type infinity() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__half>(uint16_t(0x7c00u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type quiet_NaN() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__half>(uint16_t(0x7e00u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type signaling_NaN() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__half>(uint16_t(0x7d00u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type denorm_min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__half>(uint16_t(0x0001u));
+  }
+
+  static constexpr bool is_iec559  = true;
+  static constexpr bool is_bounded = true;
+  static constexpr bool is_modulo  = false;
+
+  static constexpr bool traps                    = false;
+  static constexpr bool tinyness_before          = false;
+  static constexpr float_round_style round_style = round_to_nearest;
+};
+#endif // _CCCL_HAS_NVFP16
+
+// nvbf16
+
+#if _CCCL_HAS_NVBF16()
+template <>
+class __numeric_limits_impl<__nv_bfloat16, __numeric_limits_type::__floating_point>
+{
+public:
+  using type = __nv_bfloat16;
+
+  static constexpr bool is_specialized = true;
+
+  static constexpr bool is_signed   = true;
+  static constexpr int digits       = 8;
+  static constexpr int digits10     = 2;
+  static constexpr int max_digits10 = 4;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_bfloat16>(uint16_t(0x0080u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type max() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_bfloat16>(uint16_t(0x7f7fu));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type lowest() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_bfloat16>(uint16_t(0xff7fu));
+  }
+
+  static constexpr bool is_integer = false;
+  static constexpr bool is_exact   = false;
+  static constexpr int radix       = FLT_RADIX;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type epsilon() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_bfloat16>(uint16_t(0x3c00u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type round_error() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_bfloat16>(uint16_t(0x3f00u));
+  }
+
+  static constexpr int min_exponent   = -125;
+  static constexpr int min_exponent10 = -37;
+  static constexpr int max_exponent   = 128;
+  static constexpr int max_exponent10 = 38;
+
+  static constexpr bool has_infinity                                             = true;
+  static constexpr bool has_quiet_NaN                                            = true;
+  static constexpr bool has_signaling_NaN                                        = true;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr float_denorm_style has_denorm = denorm_present;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr bool has_denorm_loss          = false;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type infinity() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_bfloat16>(uint16_t(0x7f80u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type quiet_NaN() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_bfloat16>(uint16_t(0x7fc0u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type signaling_NaN() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_bfloat16>(uint16_t(0x7fa0u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type denorm_min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_bfloat16>(uint16_t(0x0001u));
+  }
+
+  static constexpr bool is_iec559  = true;
+  static constexpr bool is_bounded = true;
+  static constexpr bool is_modulo  = false;
+
+  static constexpr bool traps                    = false;
+  static constexpr bool tinyness_before          = false;
+  static constexpr float_round_style round_style = round_to_nearest;
+};
+#endif // _CCCL_HAS_NVBF16
+
+// nvfp8_e4m3
+
+#if _CCCL_HAS_NVFP8_E4M3()
+template <>
+class __numeric_limits_impl<__nv_fp8_e4m3, __numeric_limits_type::__floating_point>
+{
+public:
+  using type = __nv_fp8_e4m3;
+
+  static constexpr bool is_specialized = true;
+
+  static constexpr bool is_signed   = true;
+  static constexpr int digits       = 4;
+  static constexpr int digits10     = 0;
+  static constexpr int max_digits10 = 3;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e4m3>(uint8_t(0x08u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type max() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e4m3>(uint8_t(0x7eu));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type lowest() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e4m3>(uint8_t(0xfeu));
+  }
+
+  static constexpr bool is_integer = false;
+  static constexpr bool is_exact   = false;
+  static constexpr int radix       = FLT_RADIX;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type epsilon() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e4m3>(uint8_t(0x20u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type round_error() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e4m3>(uint8_t(0x30u));
+  }
+
+  static constexpr int min_exponent   = -6;
+  static constexpr int min_exponent10 = -2;
+  static constexpr int max_exponent   = 8;
+  static constexpr int max_exponent10 = 2;
+
+  static constexpr bool has_infinity                                             = false;
+  static constexpr bool has_quiet_NaN                                            = true;
+  static constexpr bool has_signaling_NaN                                        = false;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr float_denorm_style has_denorm = denorm_present;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr bool has_denorm_loss          = false;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type infinity() noexcept
+  {
+    return type{};
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type quiet_NaN() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e4m3>(uint8_t(0x7fu));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type signaling_NaN() noexcept
+  {
+    return type{};
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type denorm_min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e4m3>(uint8_t(0x01u));
+  }
+
+  static constexpr bool is_iec559  = false;
+  static constexpr bool is_bounded = true;
+  static constexpr bool is_modulo  = false;
+
+  static constexpr bool traps                    = false;
+  static constexpr bool tinyness_before          = false;
+  static constexpr float_round_style round_style = round_to_nearest;
+};
+#endif // _CCCL_HAS_NVFP8_E4M3
+
+// nvfp8_e5m2
+
+#if _CCCL_HAS_NVFP8_E5M2()
+template <>
+class __numeric_limits_impl<__nv_fp8_e5m2, __numeric_limits_type::__floating_point>
+{
+public:
+  using type = __nv_fp8_e5m2;
+
+  static constexpr bool is_specialized = true;
+
+  static constexpr bool is_signed   = true;
+  static constexpr int digits       = 3;
+  static constexpr int digits10     = 0;
+  static constexpr int max_digits10 = 2;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e5m2>(uint8_t(0x04u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type max() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e5m2>(uint8_t(0x7bu));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type lowest() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e5m2>(uint8_t(0xfbu));
+  }
+
+  static constexpr bool is_integer = false;
+  static constexpr bool is_exact   = false;
+  static constexpr int radix       = FLT_RADIX;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type epsilon() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e5m2>(uint8_t(0x34u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type round_error() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e5m2>(uint8_t(0x38u));
+  }
+
+  static constexpr int min_exponent   = -15;
+  static constexpr int min_exponent10 = -5;
+  static constexpr int max_exponent   = 15;
+  static constexpr int max_exponent10 = 4;
+
+  static constexpr bool has_infinity                                             = true;
+  static constexpr bool has_quiet_NaN                                            = true;
+  static constexpr bool has_signaling_NaN                                        = true;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr float_denorm_style has_denorm = denorm_present;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr bool has_denorm_loss          = false;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type infinity() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e5m2>(uint8_t(0x7cu));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type quiet_NaN() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e5m2>(uint8_t(0x7eu));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type signaling_NaN() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e5m2>(uint8_t(0x7du));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type denorm_min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e5m2>(uint8_t(0x01u));
+  }
+
+  static constexpr bool is_iec559  = false;
+  static constexpr bool is_bounded = true;
+  static constexpr bool is_modulo  = false;
+
+  static constexpr bool traps                    = false;
+  static constexpr bool tinyness_before          = false;
+  static constexpr float_round_style round_style = round_to_nearest;
+};
+#endif // _CCCL_HAS_NVFP8_E5M2
+
+// nvfp8_e8m0
+
+#if _CCCL_HAS_NVFP8_E8M0()
+template <>
+class __numeric_limits_impl<__nv_fp8_e8m0, __numeric_limits_type::__floating_point>
+{
+public:
+  using type = __nv_fp8_e8m0;
+
+  static constexpr bool is_specialized = true;
+
+  static constexpr bool is_signed   = false;
+  static constexpr int digits       = 1;
+  static constexpr int digits10     = 0;
+  static constexpr int max_digits10 = 2;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e8m0>(uint8_t(0x00u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type max() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e8m0>(uint8_t(0xfeu));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type lowest() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e8m0>(uint8_t(0x00u));
+  }
+
+  static constexpr bool is_integer = false;
+  static constexpr bool is_exact   = false;
+  static constexpr int radix       = FLT_RADIX;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type epsilon() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e8m0>(uint8_t(0x7fu));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type round_error() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e8m0>(uint8_t(0x7fu));
+  }
+
+  static constexpr int min_exponent   = -127;
+  static constexpr int min_exponent10 = -39;
+  static constexpr int max_exponent   = 127;
+  static constexpr int max_exponent10 = 38;
+
+  static constexpr bool has_infinity                                             = false;
+  static constexpr bool has_quiet_NaN                                            = true;
+  static constexpr bool has_signaling_NaN                                        = false;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr float_denorm_style has_denorm = denorm_absent;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr bool has_denorm_loss          = false;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type infinity() noexcept
+  {
+    return type{};
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type quiet_NaN() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp8_e8m0>(uint8_t(0xffu));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type signaling_NaN() noexcept
+  {
+    return type{};
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type denorm_min() noexcept
+  {
+    return type{};
+  }
+
+  static constexpr bool is_iec559  = false;
+  static constexpr bool is_bounded = true;
+  static constexpr bool is_modulo  = false;
+
+  static constexpr bool traps                    = false;
+  static constexpr bool tinyness_before          = false;
+  static constexpr float_round_style round_style = round_toward_zero;
+};
+#endif // _CCCL_HAS_NVFP8_E8M0()
+
+// nvfp6_e2m3
+
+#if _CCCL_HAS_NVFP6_E2M3()
+template <>
+class __numeric_limits_impl<__nv_fp6_e2m3, __numeric_limits_type::__floating_point>
+{
+public:
+  using type = __nv_fp6_e2m3;
+
+  static constexpr bool is_specialized = true;
+
+  static constexpr bool is_signed   = true;
+  static constexpr int digits       = 4;
+  static constexpr int digits10     = 0;
+  static constexpr int max_digits10 = 3;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp6_e2m3>(uint8_t(0x08u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type max() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp6_e2m3>(uint8_t(0x1fu));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type lowest() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp6_e2m3>(uint8_t(0x3fu));
+  }
+
+  static constexpr bool is_integer = false;
+  static constexpr bool is_exact   = false;
+  static constexpr int radix       = FLT_RADIX;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type epsilon() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp6_e2m3>(uint8_t(0x01u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type round_error() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp6_e2m3>(uint8_t(0x04u));
+  }
+
+  static constexpr int min_exponent   = 0;
+  static constexpr int min_exponent10 = 0;
+  static constexpr int max_exponent   = 2;
+  static constexpr int max_exponent10 = 0;
+
+  static constexpr bool has_infinity                                             = false;
+  static constexpr bool has_quiet_NaN                                            = false;
+  static constexpr bool has_signaling_NaN                                        = false;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr float_denorm_style has_denorm = denorm_present;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr bool has_denorm_loss          = false;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type infinity() noexcept
+  {
+    return type{};
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type quiet_NaN() noexcept
+  {
+    return type{};
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type signaling_NaN() noexcept
+  {
+    return type{};
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type denorm_min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp6_e2m3>(uint8_t(0x01u));
+  }
+
+  static constexpr bool is_iec559  = false;
+  static constexpr bool is_bounded = true;
+  static constexpr bool is_modulo  = false;
+
+  static constexpr bool traps                    = false;
+  static constexpr bool tinyness_before          = false;
+  static constexpr float_round_style round_style = round_to_nearest;
+};
+#endif // _CCCL_HAS_NVFP6_E2M3
+
+// nvfp6_e3m2
+
+#if _CCCL_HAS_NVFP6_E3M2()
+template <>
+class __numeric_limits_impl<__nv_fp6_e3m2, __numeric_limits_type::__floating_point>
+{
+public:
+  using type = __nv_fp6_e3m2;
+
+  static constexpr bool is_specialized = true;
+
+  static constexpr bool is_signed   = true;
+  static constexpr int digits       = 3;
+  static constexpr int digits10     = 0;
+  static constexpr int max_digits10 = 2;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp6_e3m2>(uint8_t(0x04u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type max() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp6_e3m2>(uint8_t(0x1fu));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type lowest() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp6_e3m2>(uint8_t(0x3fu));
+  }
+
+  static constexpr bool is_integer = false;
+  static constexpr bool is_exact   = false;
+  static constexpr int radix       = FLT_RADIX;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type epsilon() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp6_e3m2>(uint8_t(0x04u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type round_error() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp6_e3m2>(uint8_t(0x08u));
+  }
+
+  static constexpr int min_exponent   = -2;
+  static constexpr int min_exponent10 = -1;
+  static constexpr int max_exponent   = 4;
+  static constexpr int max_exponent10 = 1;
+
+  static constexpr bool has_infinity                                             = false;
+  static constexpr bool has_quiet_NaN                                            = false;
+  static constexpr bool has_signaling_NaN                                        = false;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr float_denorm_style has_denorm = denorm_present;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr bool has_denorm_loss          = false;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type infinity() noexcept
+  {
+    return type{};
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type quiet_NaN() noexcept
+  {
+    return type{};
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type signaling_NaN() noexcept
+  {
+    return type{};
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type denorm_min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp6_e3m2>(uint8_t(0x01u));
+  }
+
+  static constexpr bool is_iec559  = false;
+  static constexpr bool is_bounded = true;
+  static constexpr bool is_modulo  = false;
+
+  static constexpr bool traps                    = false;
+  static constexpr bool tinyness_before          = false;
+  static constexpr float_round_style round_style = round_to_nearest;
+};
+#endif // _CCCL_HAS_NVFP6_E3M2()
+
+// nvfp4_e2m1
+
+#if _CCCL_HAS_NVFP4_E2M1()
+template <>
+class __numeric_limits_impl<__nv_fp4_e2m1, __numeric_limits_type::__floating_point>
+{
+public:
+  using type = __nv_fp4_e2m1;
+
+  static constexpr bool is_specialized = true;
+
+  static constexpr bool is_signed   = true;
+  static constexpr int digits       = 2;
+  static constexpr int digits10     = 0;
+  static constexpr int max_digits10 = 2;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp4_e2m1>(uint8_t(0x2u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type max() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp4_e2m1>(uint8_t(0x7u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type lowest() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp4_e2m1>(uint8_t(0xfu));
+  }
+
+  static constexpr bool is_integer = false;
+  static constexpr bool is_exact   = false;
+  static constexpr int radix       = FLT_RADIX;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type epsilon() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp4_e2m1>(uint8_t(0x1u));
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type round_error() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp4_e2m1>(uint8_t(0x1u));
+  }
+
+  static constexpr int min_exponent   = 0;
+  static constexpr int min_exponent10 = 0;
+  static constexpr int max_exponent   = 2;
+  static constexpr int max_exponent10 = 0;
+
+  static constexpr bool has_infinity                                             = false;
+  static constexpr bool has_quiet_NaN                                            = false;
+  static constexpr bool has_signaling_NaN                                        = false;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr float_denorm_style has_denorm = denorm_present;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr bool has_denorm_loss          = false;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type infinity() noexcept
+  {
+    return type{};
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type quiet_NaN() noexcept
+  {
+    return type{};
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type signaling_NaN() noexcept
+  {
+    return type{};
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type denorm_min() noexcept
+  {
+    return _CUDA_VSTD::__fp_from_storage<__nv_fp4_e2m1>(uint8_t(0x1u));
+  }
+
+  static constexpr bool is_iec559  = false;
+  static constexpr bool is_bounded = true;
+  static constexpr bool is_modulo  = false;
+
+  static constexpr bool traps                    = false;
+  static constexpr bool tinyness_before          = false;
+  static constexpr float_round_style round_style = round_to_nearest;
+};
+#endif // _CCCL_HAS_NVFP4_E2M1()
+
+// __float128
+
+#if _CCCL_HAS_FLOAT128()
+template <>
+class __numeric_limits_impl<__float128, __numeric_limits_type::__floating_point>
+{
+public:
+  using type = __float128;
+
+  static constexpr bool is_specialized = true;
+
+  static constexpr bool is_signed   = true;
+  static constexpr int digits       = 113;
+  static constexpr int digits10     = 33;
+  static constexpr int max_digits10 = 36;
+
+#  if _CCCL_HAS_FLOAT128_LITERAL()
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type min() noexcept
+  {
+    return _CCCL_FLOAT128_LITERAL(3.36210314311209350626267781732175260e-4932);
+  }
+#  else // ^^^ _CCCL_HAS_FLOAT128_LITERAL() ^^^ // vvv !_CCCL_HAS_FLOAT128_LITERAL() vvv
+  _LIBCUDACXX_HIDE_FROM_ABI static _LIBCUDACXX_CONSTEXPR_BIT_CAST type min() noexcept
+  {
+    return _CUDA_VSTD::bit_cast<type>(__uint128_t{0x0001'0000'0000'0000} << 64);
+  }
+#  endif // ^^^ !_CCCL_HAS_FLOAT128_LITERAL() ^^^
+#  if _CCCL_HAS_FLOAT128_LITERAL()
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type max() noexcept
+  {
+    return _CCCL_FLOAT128_LITERAL(1.18973149535723176508575932662800702e+4932);
+  }
+#  else // ^^^ _CCCL_HAS_FLOAT128_LITERAL() ^^^ // vvv !_CCCL_HAS_FLOAT128_LITERAL() vvv
+  _LIBCUDACXX_HIDE_FROM_ABI static _LIBCUDACXX_CONSTEXPR_BIT_CAST type max() noexcept
+  {
+    return _CUDA_VSTD::bit_cast<type>(__uint128_t{0x7ffe'ffff'ffff'ffff} << 64 | 0xffff'ffff'ffff'ffff);
+  }
+#  endif // ^^^ !_CCCL_HAS_FLOAT128_LITERAL() ^^^
+#  if _CCCL_HAS_FLOAT128_LITERAL()
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type lowest() noexcept
+  {
+    return -max();
+  }
+#  else // ^^^ _CCCL_HAS_FLOAT128_LITERAL() ^^^ // vvv !_CCCL_HAS_FLOAT128_LITERAL() vvv
+  _LIBCUDACXX_HIDE_FROM_ABI static _LIBCUDACXX_CONSTEXPR_BIT_CAST type lowest() noexcept
+  {
+    return -max();
+  }
+#  endif // ^^^ !_CCCL_HAS_FLOAT128_LITERAL() ^^^
+
+  static constexpr bool is_integer = false;
+  static constexpr bool is_exact   = false;
+  static constexpr int radix       = FLT_RADIX;
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type epsilon() noexcept
+  {
+    return type(1.92592994438723585305597794258492732e-34);
+  }
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type round_error() noexcept
+  {
+    return type(0.5);
+  }
+
+  static constexpr int min_exponent   = -16381;
+  static constexpr int min_exponent10 = -4931;
+  static constexpr int max_exponent   = 16384;
+  static constexpr int max_exponent10 = 4932;
+
+  static constexpr bool has_infinity                                             = true;
+  static constexpr bool has_quiet_NaN                                            = true;
+  static constexpr bool has_signaling_NaN                                        = true;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr float_denorm_style has_denorm = denorm_present;
+  _LIBCUDACXX_DEPRECATED_IN_CXX23 static constexpr bool has_denorm_loss          = false;
+
+#  if defined(_CCCL_BUILTIN_HUGE_VALF128)
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type infinity() noexcept
+  {
+    return _CCCL_BUILTIN_HUGE_VALF128();
+  }
+#  else // ^^^ _CCCL_BUILTIN_HUGE_VALF128 ^^^ // vvv !_CCCL_BUILTIN_HUGE_VALF128 vvv
+  _LIBCUDACXX_HIDE_FROM_ABI static _LIBCUDACXX_CONSTEXPR_BIT_CAST type infinity() noexcept
+  {
+    return _CUDA_VSTD::bit_cast<type>(__uint128_t{0x7fff'0000'0000'0000} << 64);
+  }
+#  endif // ^^^ !_CCCL_BUILTIN_HUGE_VALF128 ^^^
+#  if defined(_CCCL_BUILTIN_NANF128)
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type quiet_NaN() noexcept
+  {
+    return _CCCL_BUILTIN_NANF128("");
+  }
+#  else // ^^^ _CCCL_BUILTIN_NANF128 ^^^ // vvv !_CCCL_BUILTIN_NANF128 vvv
+  _LIBCUDACXX_HIDE_FROM_ABI static _LIBCUDACXX_CONSTEXPR_BIT_CAST type quiet_NaN() noexcept
+  {
+    return _CUDA_VSTD::bit_cast<type>(__uint128_t{0x7fff'8000'0000'0000} << 64);
+  }
+#  endif // ^^^ !_CCCL_BUILTIN_NANF128 ^^^
+#  if defined(_CCCL_BUILTIN_NANSF128)
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type signaling_NaN() noexcept
+  {
+    return _CCCL_BUILTIN_NANSF128("");
+  }
+#  else // ^^^ _CCCL_BUILTIN_NANSF128 ^^^ // vvv !_CCCL_BUILTIN_NANSF128 vvv
+  _LIBCUDACXX_HIDE_FROM_ABI static _LIBCUDACXX_CONSTEXPR_BIT_CAST type signaling_NaN() noexcept
+  {
+    return _CUDA_VSTD::bit_cast<type>(__uint128_t{0x7fff'4000'0000'0000} << 64);
+  }
+#  endif // ^^^ !_CCCL_BUILTIN_NANSF128 ^^^
+#  if _CCCL_HAS_FLOAT128_LITERAL()
+  _LIBCUDACXX_HIDE_FROM_ABI static constexpr type denorm_min() noexcept
+  {
+    return _CCCL_FLOAT128_LITERAL(6.47517511943802511092443895822764655e-4966);
+  }
+#  else // ^^^ _CCCL_HAS_FLOAT128_LITERAL() ^^^ // vvv !_CCCL_HAS_FLOAT128_LITERAL() vvv
+  _LIBCUDACXX_HIDE_FROM_ABI static _LIBCUDACXX_CONSTEXPR_BIT_CAST type denorm_min() noexcept
+  {
+    return _CUDA_VSTD::bit_cast<type>(__uint128_t{0x1});
+  }
+#  endif // ^^^ !_CCCL_HAS_FLOAT128_LITERAL() ^^^
+
+  static constexpr bool is_iec559  = true;
+  static constexpr bool is_bounded = true;
+  static constexpr bool is_modulo  = false;
+
+  static constexpr bool traps                    = false;
+  static constexpr bool tinyness_before          = false;
+  static constexpr float_round_style round_style = round_to_nearest;
+};
+#endif // _CCCL_HAS_FLOAT128()
+
+_LIBCUDACXX_END_NAMESPACE_STD
+
+#include <cuda/std/__cccl/epilogue.h>
+
+#endif // _LIBCUDACXX___LIMITS_NUMERIC_LIMITS_EXT_H