module;

#include <type_traits>
#include <concepts>
#include <immintrin.h>
#include <string>
#include <sstream>
#include <iostream>

export module Crafter.Math:Vector;

import :BasicTypes;

namespace Crafter {
	export template <typename T, uint32_t len>
		class Vector {
		typedef
			typename std::conditional<(sizeof(T)* len > 32 && (std::same_as<T, int64_t> || std::same_as<T, int32_t> || std::same_as<T, int16_t> || std::same_as<T, int8_t> || std::same_as<T, uint64_t> || std::same_as<T, uint32_t> || std::same_as<T, uint16_t> || std::same_as<T, uint8_t>)), __m512i,
			typename std::conditional<(sizeof(T)* len > 16 && (std::same_as<T, int64_t> || std::same_as<T, int32_t> || std::same_as<T, int16_t> || std::same_as<T, int8_t> || std::same_as<T, uint64_t> || std::same_as<T, uint32_t> || std::same_as<T, uint16_t> || std::same_as<T, uint8_t>)), __m256i,
			typename std::conditional<(sizeof(T)* len <= 16 && (std::same_as<T, int64_t> || std::same_as<T, int32_t> || std::same_as<T, int16_t> || std::same_as<T, int8_t> || std::same_as<T, uint64_t> || std::same_as<T, uint32_t> || std::same_as<T, uint16_t> || std::same_as<T, uint8_t>)), __m128i,
			typename std::conditional<(len > 16 && std::same_as<T, __fp16>), __m512h,
			typename std::conditional<(len > 8 && std::same_as<T, __fp16>), __m256h,
			typename std::conditional<(len <= 8 && std::same_as<T, __fp16>), __m128h,
			typename std::conditional<(len > 8 && std::same_as<T, float>), __m512,
			typename std::conditional<(len > 4 && std::same_as<T, float>), __m256,
			typename std::conditional<(len <= 4 && std::same_as<T, float>), __m128,
			typename std::conditional<(len > 4 && std::same_as<T, double>), __m512d,
			typename std::conditional<(len > 2 && std::same_as<T, double>), __m256d, __m128d
			>::type>::type>::type>::type>::type>::type>::type>::type>::type>::type>::type vector_type;


		public:
		template <typename Datatype, typename Vectortype>
		static consteval uint8_t GetVectorAlignedSize() {
			if constexpr(std::same_as<Datatype, int8_t> && std::same_as<Vectortype, __m512i>) {
				return 64;
			} else if constexpr(std::same_as<Vectortype, __m512h> || (std::same_as<T, int8_t> && std::same_as<Vectortype, __m256i>) || (std::same_as<T, int16_t> && std::same_as<Vectortype, __m512i>)) {
				return 32;
			} else if constexpr(std::same_as<Vectortype, __m256h> || std::same_as<Vectortype, __m512> || (std::same_as<Datatype, int8_t> && std::same_as<Vectortype, __m128i>) || (std::same_as<Datatype, int16_t> && std::same_as<Vectortype, __m256i>) || (std::same_as<T, int32_t> && std::same_as<Vectortype, __m512i>)) {
				return 16;
			} else if constexpr(std::same_as<Vectortype, __m128h> || std::same_as<Vectortype, __m256> || std::same_as<Vectortype, __m512d> || (std::same_as<Datatype, int16_t> && std::same_as<Vectortype, __m128i>) || (std::same_as<T, int32_t> && std::same_as<Vectortype, __m256i>) || (std::same_as<Datatype, int64_t> && std::same_as<Vectortype, __m512i>)) {
				return 8;
			} else if constexpr(std::same_as<Vectortype, __m128> || std::same_as<Vectortype, __m256d> || (std::same_as<Datatype, int32_t> && std::same_as<Vectortype, __m128>) || (std::same_as<Datatype, int64_t> && std::same_as<Vectortype, __m256i>)) {
				return 4;
			} else if constexpr(std::same_as<Vectortype, __m128d> || (std::same_as<Datatype, int64_t> && std::same_as<Vectortype, __m128>)) {
				return 2;
			} else{
				throw std::invalid_argument("");
			}
		}
			vector_type v;

			Vector() {};
			Vector(__m128h v) requires(std::same_as<vector_type, __m128h>) : v(v) { }
			Vector(__m128 v) requires(std::same_as<vector_type, __m128>) : v(v) { }
			Vector(__m128d v) requires(std::same_as<vector_type, __m128d>) : v(v) { }
			Vector(__m128i v) requires(std::same_as<vector_type, __m128i>) : v(v) { }

			Vector(__m256h v) requires(std::same_as<vector_type, __m256h>) : v(v) { }
			Vector(__m256 v) requires(std::same_as<vector_type, __m256>) : v(v) { }
			Vector(__m256d v) requires(std::same_as<vector_type, __m256d>) : v(v) { }
			Vector(__m256i v) requires(std::same_as<vector_type, __m256i>) : v(v) { }

			Vector(__m512h v) requires(std::same_as<vector_type, __m512h>) : v(v) { }
			Vector(__m512 v) requires(std::same_as<vector_type, __m512>) : v(v) { }
			Vector(__m512d v) requires(std::same_as<vector_type, __m512d>) : v(v) { }
			Vector(__m512i v) requires(std::same_as<vector_type, __m512i>) : v(v) { }

			template <uint32_t blen>
			void operator+=(Vector<T, blen> b) requires(Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				this->v = (*this+b).v;
			}
			template <uint32_t blen>
			void operator-=(Vector<T, blen> b) requires(Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				this->v = (*this-b).v;
			}
			template <uint32_t blen>
			void operator*=(Vector<T, blen> b) requires(Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				this->v = (*this*b).v;
			}
			template <uint32_t blen>
			void operator/=(Vector<T, blen> b) requires(Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				this->v = (*this/b).v;
			}

			std::string ToString() const {
				std::ostringstream ss;
				ss << "{ ";
				T store[GetVectorAlignedSize<T, vector_type>()];
				Store(store);
				for(uint8_t i = 0; i < len; i++) {
					ss << std::format("{}", store[i]);
					if(i+1 < len) {
						ss << ", ";
					}
				}
				ss << " }";
				return std::string(ss.str());
			}
#pragma region 128
			Vector(
				const __fp16& x0 = 0, const __fp16& y0 = 0, const __fp16& z0 = 0, const __fp16& w0 = 0,
				const __fp16& x1 = 0, const __fp16& y1 = 0, const __fp16& z1 = 0, const __fp16& w1 = 0
			) requires(std::same_as<T, __fp16> && std::same_as<vector_type, __m128h>) {
				__fp16 temp[]{ x0,y0,z0,w0,x1,y1,z1,w1,};
				v = _mm_load_ph(temp);
			}

			Vector(float x0 = 0, float y0 = 0, float z0 = 0, float w0 = 0) requires(std::same_as<T, float>&& std::same_as<vector_type, __m128>) {
				v = _mm_set_ps(w0, z0, y0, x0);
			}

			Vector(double x0 = 0, double y0 = 0) requires(std::same_as<T, double>&& std::same_as<vector_type, __m128d>) {
				v = _mm_set_pd(y0, x0);
			}

			Vector(
				int8_t x0 = 0, int8_t y0 = 0, int8_t z0 = 0, int8_t w0 = 0,
				int8_t x1 = 0, int8_t y1 = 0, int8_t z1 = 0, int8_t w1 = 0,
				int8_t x2 = 0, int8_t y2 = 0, int8_t z2 = 0, int8_t w2 = 0,
				int8_t x3 = 0, int8_t y3 = 0, int8_t z3 = 0, int8_t w3 = 0
			) requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m128i>) {
				v = _mm_set_epi8(w3, z3, y3, x3, w2, z2, y2, x2, w1, z1, y1, x1, w0, z0, y0, x0);
			}

			Vector(
				int16_t x0 = 0, int16_t y0 = 0, int16_t z0 = 0, int16_t w0 = 0,
				int16_t x1 = 0, int16_t y1 = 0, int16_t z1 = 0, int16_t w1 = 0
			) requires(std::same_as<T, int16_t>&& std::same_as<vector_type, __m128i>) {
				v = _mm_set_epi16(w1, z1, y1, x1, w0, z0, y0, x0);
			}

			Vector(int32_t x0 = 0, int32_t y0 = 0, int32_t z0 = 0, int32_t w0 = 0) requires(std::same_as<T, int32_t>&& std::same_as<vector_type, __m128i>) {
				v = _mm_set_epi32(w0, z0, y0, x0);
			}

			Vector(int64_t x0 = 0, int64_t y0 = 0) requires(std::same_as<T, int64_t>&& std::same_as<vector_type, __m128i>) {
				v = _mm_set_epi64x(y0, x0);
			}

			Vector(
				uint8_t x0 = 0, uint8_t y0 = 0, uint8_t z0 = 0, uint8_t w0 = 0,
				uint8_t x1 = 0, uint8_t y1 = 0, uint8_t z1 = 0, uint8_t w1 = 0,
				uint8_t x2 = 0, uint8_t y2 = 0, uint8_t z2 = 0, uint8_t w2 = 0,
				uint8_t x3 = 0, uint8_t y3 = 0, uint8_t z3 = 0, uint8_t w3 = 0
			) requires(std::same_as<T, uint8_t> && std::same_as<vector_type, __m128i>) {
				v = _mm_set_epi8(w3, z3, y3, x3, w2, z2, y2, x2, w1, z1, y1, x1, w0, z0, y0, x0);
			}

			Vector(
				uint16_t x0 = 0, uint16_t y0 = 0, uint16_t z0 = 0, uint16_t w0 = 0,
				uint16_t x1 = 0, uint16_t y1 = 0, uint16_t z1 = 0, uint16_t w1 = 0
			) requires(std::same_as<T, uint16_t>&& std::same_as<vector_type, __m128i>) {
				v = _mm_set_epi16(w1, z1, y1, x1, w0, z0, y0, x0);
			}

			Vector(uint32_t x0 = 0, uint32_t y0 = 0, uint32_t z0 = 0, uint32_t w0 = 0) requires(std::same_as<T, uint32_t>&& std::same_as<vector_type, __m128i>) {
				v = _mm_set_epi32(w0, z0, y0, x0);
			}

			Vector(uint64_t x0 = 0, uint64_t y0 = 0) requires(std::same_as<T, uint64_t>&& std::same_as<vector_type, __m128i>) {
				v = _mm_set_epi64x(y0, x0);
			}

			static Vector<T, len> Zero() requires(std::same_as<vector_type, __m128>) {
				return Vector<T, len>(_mm_setzero_ps());
			}

			void Store(T* data) const requires(std::same_as<vector_type, __m128h>) {
				_mm_storeu_ph(reinterpret_cast<void*>(data), reinterpret_cast<__m128h>(v));
			}
			void Store(T* data) const requires(std::same_as<vector_type, __m128>) {
				_mm_storeu_ps(reinterpret_cast<float*>(data), reinterpret_cast<__m128>(v));
			}
			void Store(T* data) const requires(std::same_as<vector_type, __m128d>) {
				_mm_storeu_pd(data, reinterpret_cast<__m128d>(v));
			}
			void Store(T* data) const requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m128i>) {
				_mm_storeu_epi8(reinterpret_cast<void*>(data), reinterpret_cast<__m128i>(v));
			}
			void Store(T* data) const requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m128i>) {
				_mm_storeu_epi16(reinterpret_cast<void*>(data), reinterpret_cast<__m128i>(v));
			}
			void Store(T* data) const requires(std::same_as<T, int32_t> && std::same_as<vector_type, __m128i>) {
				_mm_storeu_epi32(reinterpret_cast<void*>(data), reinterpret_cast<__m128i>(v));
			}
			void Store(T* data) const requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m128i>) {
				_mm_storeu_epi64(reinterpret_cast<void*>(data), reinterpret_cast<__m128i>(v));
			}

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<vector_type, __m128h> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_add_ph(reinterpret_cast<__m128h>(v), reinterpret_cast<__m128h>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<vector_type, __m128h> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_sub_ph(reinterpret_cast<__m128h>(v), reinterpret_cast<__m128h>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<vector_type, __m128h> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_mul_ph(reinterpret_cast<__m128h>(v), reinterpret_cast<__m128h>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<vector_type, __m128h> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_div_ph(reinterpret_cast<__m128h>(v), reinterpret_cast<__m128h>(b.v)));
			}

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<vector_type, __m128> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_add_ps(reinterpret_cast<__m128>(v), reinterpret_cast<__m128>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<vector_type, __m128> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_sub_ps(reinterpret_cast<__m128>(v), reinterpret_cast<__m128>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<vector_type, __m128> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_mul_ps(reinterpret_cast<__m128>(v), reinterpret_cast<__m128>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<vector_type, __m128> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_div_ps(reinterpret_cast<__m128>(v), reinterpret_cast<__m128>(b.v)));
			}

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<vector_type, __m128d> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_add_pd(reinterpret_cast<__m128d>(v), reinterpret_cast<__m128d>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<vector_type, __m128d> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_sub_pd(reinterpret_cast<__m128d>(v), reinterpret_cast<__m128d>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<vector_type, __m128d> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_mul_pd(reinterpret_cast<__m128d>(v), reinterpret_cast<__m128d>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<vector_type, __m128d> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_div_pd(reinterpret_cast<__m128d>(v), reinterpret_cast<__m128d>(b.v)));
			}

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m128i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_add_epi8(reinterpret_cast<__m128i>(v), reinterpret_cast<__m128i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m128i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_sub_epi8(reinterpret_cast<__m128i>(v), reinterpret_cast<__m128i>(b.v)));
			}
			// template <uint32_t blen>
			// Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_mul_epi8(v, bv));
			// }
			// template <uint32_t blen>
			// Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_div_epi8(v, bv));
			// }

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m128i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_add_epi16(reinterpret_cast<__m128i>(v), reinterpret_cast<__m128i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m128i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_sub_epi16(reinterpret_cast<__m128i>(v), reinterpret_cast<__m128i>(b.v)));
			}
			// template <uint32_t blen>
			// Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_mul_epi16(v, bv));
			// }
			// template <uint32_t blen>
			// Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_div_epi16(v, bv));
			// }

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<T, int32_t> && std::same_as<vector_type, __m128i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_add_epi32(reinterpret_cast<__m128i>(v), reinterpret_cast<__m128i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<T, int32_t> && std::same_as<vector_type, __m128i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_sub_epi32(reinterpret_cast<__m128i>(v), reinterpret_cast<__m128i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<T, int32_t> && std::same_as<vector_type, __m128i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_mul_epi32(reinterpret_cast<__m128i>(v), reinterpret_cast<__m128i>(b.v)));
			}
			// template <uint32_t blen>
			// Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<T, i132> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_div_epi32(v, bv));
			// }

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m128i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_add_epi64(reinterpret_cast<__m128i>(v), reinterpret_cast<__m128i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m128i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm_sub_epi64(reinterpret_cast<__m128i>(v), reinterpret_cast<__m128i>(b.v)));
			}
			// template <uint32_t blen>
			// Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_mul_epi64(v, bv));
			// }
			// template <uint32_t blen>
			// Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_div_epi64(v, bv));
			// }

			template <uint32_t blen>
			Vector<T, len> AddMask() {

			}
#pragma endregion
#pragma region 256
			Vector(
				const __fp16& x0 = 0, const __fp16& y0 = 0, const __fp16& z0 = 0, const __fp16& w0 = 0,
				const __fp16& x1 = 0, const __fp16& y1 = 0, const __fp16& z1 = 0, const __fp16& w1 = 0,
				const __fp16& x2 = 0, const __fp16& y2 = 0, const __fp16& z2 = 0, const __fp16& w2 = 0,
				const __fp16& x3 = 0, const __fp16& y3 = 0, const __fp16& z3 = 0, const __fp16& w3 = 0
			) requires(std::same_as<T, __fp16>&& std::same_as<vector_type, __m256h>) {
				__fp16 temp[]{ w0,z0,y0,x0,w1,z1,y1,x1,w2,z2,y2,x2,w3,z3,y3,x3 };
				v = _mm256_load_ph(temp);
			}

			Vector(
				float x0 = 0, float y0 = 0, float z0 = 0, float w0 = 0,
				float x1 = 0, float y1 = 0, float z1 = 0, float w1 = 0
			) requires(std::same_as<T, float>&& std::same_as<vector_type, __m256>) {
				v = _mm256_set_ps(w1, z1, y1, x1, w0, z0, y0, x0);
			}

			Vector(double x0 = 0, double y0 = 0, double z0 = 0, double w0 = 0) requires(std::same_as<T, double>&& std::same_as<vector_type, __m256d>) {
				v = _mm256_set_pd(w0, z0, y0, x0);
			}

			Vector(
				int8_t x0 = 0, int8_t y0 = 0, int8_t z0 = 0, int8_t w0 = 0,
				int8_t x1 = 0, int8_t y1 = 0, int8_t z1 = 0, int8_t w1 = 0,
				int8_t x2 = 0, int8_t y2 = 0, int8_t z2 = 0, int8_t w2 = 0,
				int8_t x3 = 0, int8_t y3 = 0, int8_t z3 = 0, int8_t w3 = 0,
				int8_t x4 = 0, int8_t y4 = 0, int8_t z4 = 0, int8_t w4 = 0,
				int8_t x5 = 0, int8_t y5 = 0, int8_t z5 = 0, int8_t w5 = 0,
				int8_t x6 = 0, int8_t y6 = 0, int8_t z6 = 0, int8_t w6 = 0,
				int8_t x7 = 0, int8_t y7 = 0, int8_t z7 = 0, int8_t w7 = 0
			) requires(std::same_as<T, int8_t>&& std::same_as<vector_type, __m256i>) {
				v = _mm256_set_epi8(w7, z7, y7, x7, w6, z6, y6, x6, w5, z5, y5, x5, w4, z4, y4, x4, w3, z3, y3, x3, w2, z2, y2, x2, w1, z1, y1, x1, w0, z0, y0, x0);
			}

			Vector(
				int16_t x0 = 0, int16_t y0 = 0, int16_t z0 = 0, int16_t w0 = 0,
				int16_t x1 = 0, int16_t y1 = 0, int16_t z1 = 0, int16_t w1 = 0,
				int16_t x2 = 0, int16_t y2 = 0, int16_t z2 = 0, int16_t w2 = 0,
				int16_t x3 = 0, int16_t y3 = 0, int16_t z3 = 0, int16_t w3 = 0
			) requires(std::same_as<T, int16_t>&& std::same_as<vector_type, __m256i>) {
				v = _mm256_set_epi16(w3, z3, y3, x3, w2, z2, y2, x2, w1, z1, y1, x1, w0, z0, y0, x0);
			}

			Vector(
				int32_t x0 = 0, int32_t y0 = 0, int32_t z0 = 0, int32_t w0 = 0,
				int32_t x1 = 0, int32_t y1 = 0, int32_t z1 = 0, int32_t w1 = 0
			) requires(std::same_as<T, int32_t>&& std::same_as<vector_type, __m256i>) {
				v = _mm256_set_epi32(w1, z1, y1, x1, w0, z0, y0, x0);
			}

			Vector(int64_t x0 = 0, int64_t y0 = 0, int64_t z0 = 0, int64_t w0 = 0) requires(std::same_as<T, int64_t>&& std::same_as<vector_type, __m256>) {
				v = _mm256_set_epi64x(w0, z0, y0, x0);
			}

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<vector_type, __m256h> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_add_ph(reinterpret_cast<__m256h>(v), reinterpret_cast<__m256h>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<vector_type, __m256h> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_sub_ph(reinterpret_cast<__m256h>(v), reinterpret_cast<__m256h>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<vector_type, __m256h> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_mul_ph(reinterpret_cast<__m256h>(v), reinterpret_cast<__m256h>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<vector_type, __m256h> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_div_ph(reinterpret_cast<__m256h>(v), reinterpret_cast<__m256h>(b.v)));
			}

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<vector_type, __m256> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_add_ps(reinterpret_cast<__m256>(v), reinterpret_cast<__m256>(b.v)));
			}

			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<vector_type, __m256> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_sub_ps(reinterpret_cast<__m256>(v), reinterpret_cast<__m256>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<vector_type, __m256> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_mul_ps(reinterpret_cast<__m256>(v), reinterpret_cast<__m256>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<vector_type, __m256> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_div_ps(reinterpret_cast<__m256>(v), reinterpret_cast<__m256>(b.v)));
			}

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<vector_type, __m256d> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_add_pd(reinterpret_cast<__m256d>(v), reinterpret_cast<__m256d>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<vector_type, __m256d> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_sub_pd(reinterpret_cast<__m256d>(v), reinterpret_cast<__m256d>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<vector_type, __m256d> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_mul_pd(reinterpret_cast<__m256d>(v), reinterpret_cast<__m256d>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<vector_type, __m256d> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_div_pd(reinterpret_cast<__m256d>(v), reinterpret_cast<__m256d>(b.v)));
			}

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m256i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_add_epi8(reinterpret_cast<__m256i>(v), reinterpret_cast<__m256i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m256i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_sub_epi8(reinterpret_cast<__m256i>(v), reinterpret_cast<__m256i>(b.v)));
			}
			// template <uint32_t blen>
			// Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_mul_epi8(v, bv));
			// }
			// template <uint32_t blen>
			// Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_div_epi8(v, bv));
			// }

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m256i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_add_epi16(reinterpret_cast<__m256i>(v), reinterpret_cast<__m256i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m256i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_sub_epi16(reinterpret_cast<__m256i>(v), reinterpret_cast<__m256i>(b.v)));
			}
			// template <uint32_t blen>
			// Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_mul_epi16(v, bv));
			// }
			// template <uint32_t blen>
			// Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_div_epi16(v, bv));
			// }

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<T, int32_t> && std::same_as<vector_type, __m256i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_add_epi32(reinterpret_cast<__m256i>(v), reinterpret_cast<__m256i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<T, int32_t> && std::same_as<vector_type, __m256i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_sub_epi32(reinterpret_cast<__m256i>(v), reinterpret_cast<__m256i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<T, int32_t> && std::same_as<vector_type, __m256i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_mul_epi32(reinterpret_cast<__m256i>(v), reinterpret_cast<__m256i>(b.v)));
			}
			// template <uint32_t blen>
			// Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<T, i132> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_div_epi32(v, bv));
			// }

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m256i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_add_epi64(reinterpret_cast<__m256i>(v), reinterpret_cast<__m256i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m256i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm256_sub_epi64(reinterpret_cast<__m256i>(v), reinterpret_cast<__m256i>(b.v)));
			}
			// template <uint32_t blen>
			// Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_mul_epi64(v, bv));
			// }
			// template <uint32_t blen>
			// Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_div_epi64(v, bv));
			// }

			void Store(T* data) const requires(std::same_as<vector_type, __m256h>) {
				_mm256_storeu_ph(reinterpret_cast<void*>(data), reinterpret_cast<__m256h>(v));
			}
			void Store(T* data) const requires(std::same_as<vector_type, __m256>) {
				_mm256_storeu_ps(data, reinterpret_cast<__m256>(v));
			}
			void Store(T* data) const requires(std::same_as<vector_type, __m256d>) {
				_mm256_storeu_pd(data, reinterpret_cast<__m256d>(v));
			}
			void Store(T* data) const requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m256i>) {
				_mm256_storeu_epi8(reinterpret_cast<void*>(data), reinterpret_cast<__m256i>(v));
			}
			void Store(T* data) const requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m256i>) {
				_mm256_storeu_epi16(reinterpret_cast<void*>(data), reinterpret_cast<__m256i>(v));
			}
			void Store(T* data) const requires(std::same_as<T, int32_t> && std::same_as<vector_type, __m256i>) {
				_mm256_storeu_epi32(reinterpret_cast<void*>(data), reinterpret_cast<__m256i>(v));
			}
			void Store(T* data) const requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m256i>) {
				_mm256_storeu_epi64(reinterpret_cast<void*>(data), reinterpret_cast<__m256i>(v));
			}
#pragma endregion
#pragma region 512
			Vector(
				const __fp16& x0 = 0, const __fp16& y0 = 0, const __fp16& z0 = 0, const __fp16& w0 = 0,
				const __fp16& x1 = 0, const __fp16& y1 = 0, const __fp16& z1 = 0, const __fp16& w1 = 0,
				const __fp16& x2 = 0, const __fp16& y2 = 0, const __fp16& z2 = 0, const __fp16& w2 = 0,
				const __fp16& x3 = 0, const __fp16& y3 = 0, const __fp16& z3 = 0, const __fp16& w3 = 0,
				const __fp16& x4 = 0, const __fp16& y4 = 0, const __fp16& z4 = 0, const __fp16& w4 = 0,
				const __fp16& x5 = 0, const __fp16& y5 = 0, const __fp16& z5 = 0, const __fp16& w5 = 0,
				const __fp16& x6 = 0, const __fp16& y6 = 0, const __fp16& z6 = 0, const __fp16& w6 = 0,
				const __fp16& x7 = 0, const __fp16& y7 = 0, const __fp16& z7 = 0, const __fp16& w7 = 0
			) requires(std::same_as<T, __fp16>&& std::same_as<vector_type, __m512h>) {
				__fp16 temp[]{ w0,z0,y0,x0, w1,z1,y1,x1, w2,z2,y2,x2,w3, z3,y3,x3, w4,z4,y4,x4, w5,z5,y5,x5, w6,z6,y6,x6, w7,z7,y7,x7 };
				v = _mm512_load_ph(temp);
			}

			Vector(
				float x0 = 0, float y0 = 0, float z0 = 0, float w0 = 0,
				float x1 = 0, float y1 = 0, float z1 = 0, float w1 = 0,
				float x2 = 0, float y2 = 0, float z2 = 0, float w2 = 0,
				float x3 = 0, float y3 = 0, float z3 = 0, float w3 = 0
			) requires(std::same_as<T, float>&& std::same_as<vector_type, __m512>) {
				v = _mm512_set_ps(
					w3, z3, y3, x3,
					w2, z2, y2, x2,
					w1, z1, y1, x1,
					w0, z0, y0, x0
				);
			}

			Vector(
				double x0 = 0, double y0 = 0, double z0 = 0, double w0 = 0,
				double x1 = 0, double y1 = 0, double z1 = 0, double w1 = 0
			) requires(std::same_as<T, double>&& std::same_as<vector_type, __m512d>) {
				v = _mm512_set_pd(
					w1, z1, y1, x1,
					w0, z0, y0, x0
				);
			}

			Vector(
				int8_t x0 = 0, int8_t y0 = 0, int8_t z0 = 0, int8_t w0 = 0,
				int8_t x1 = 0, int8_t y1 = 0, int8_t z1 = 0, int8_t w1 = 0,
				int8_t x2 = 0, int8_t y2 = 0, int8_t z2 = 0, int8_t w2 = 0,
				int8_t x3 = 0, int8_t y3 = 0, int8_t z3 = 0, int8_t w3 = 0,
				int8_t x4 = 0, int8_t y4 = 0, int8_t z4 = 0, int8_t w4 = 0,
				int8_t x5 = 0, int8_t y5 = 0, int8_t z5 = 0, int8_t w5 = 0,
				int8_t x6 = 0, int8_t y6 = 0, int8_t z6 = 0, int8_t w6 = 0,
				int8_t x7 = 0, int8_t y7 = 0, int8_t z7 = 0, int8_t w7 = 0,
				int8_t x8 = 0, int8_t y8 = 0, int8_t z8 = 0, int8_t w8 = 0,
				int8_t x9 = 0, int8_t y9 = 0, int8_t z9 = 0, int8_t w9 = 0,
				int8_t x10 = 0, int8_t y10 = 0, int8_t z10 = 0, int8_t w10 = 0,
				int8_t x11 = 0, int8_t y11 = 0, int8_t z11 = 0, int8_t w11 = 0,
				int8_t x12 = 0, int8_t y12 = 0, int8_t z12 = 0, int8_t w12 = 0,
				int8_t x13 = 0, int8_t y13 = 0, int8_t z13 = 0, int8_t w13 = 0,
				int8_t x14 = 0, int8_t y14 = 0, int8_t z14 = 0, int8_t w14 = 0,
				int8_t x15 = 0, int8_t y15 = 0, int8_t z15 = 0, int8_t w15 = 0
			) requires(std::same_as<T, int8_t>&& std::same_as<vector_type, __m512i>) {
				v = _mm512_set_epi8(
					w15, z15, y15, x15,
					w14, z14, y14, x14,
					w13, z13, y13, x13,
					w12, z12, y12, x12,
					w11, z11, y11, x11,
					w10, z10, y10, x10,
					w9, z9, y9, x9,
					w8, z8, y8, x8,
					w7, z7, y7, x7,
					w6, z6, y6, x6,
					w5, z5, y5, x5,
					w4, z4, y4, x4,
					w3, z3, y3, x3,
					w2, z2, y2, x2,
					w1, z1, y1, x1,
					w0, z0, y0, x0
				);
			}

			Vector(
				int16_t x0 = 0, int16_t y0 = 0, int16_t z0 = 0, int16_t w0 = 0,
				int16_t x1 = 0, int16_t y1 = 0, int16_t z1 = 0, int16_t w1 = 0,
				int16_t x2 = 0, int16_t y2 = 0, int16_t z2 = 0, int16_t w2 = 0,
				int16_t x3 = 0, int16_t y3 = 0, int16_t z3 = 0, int16_t w3 = 0,
				int16_t x4 = 0, int16_t y4 = 0, int16_t z4 = 0, int16_t w4 = 0,
				int16_t x5 = 0, int16_t y5 = 0, int16_t z5 = 0, int16_t w5 = 0,
				int16_t x6 = 0, int16_t y6 = 0, int16_t z6 = 0, int16_t w6 = 0,
				int16_t x7 = 0, int16_t y7 = 0, int16_t z7 = 0, int16_t w7 = 0
			) requires(std::same_as<T, int16_t>&& std::same_as<vector_type, __m512i>) {
				v = _mm512_set_epi16(
					w7, z7, y7, x7,
					w6, z6, y6, x6,
					w5, z5, y5, x5,
					w4, z4, y4, x4,
					w3, z3, y3, x3,
					w2, z2, y2, x2,
					w1, z1, y1, x1,
					w0, z0, y0, x0
				);
			}

			Vector(
				int32_t x0 = 0, int32_t y0 = 0, int32_t z0 = 0, int32_t w0 = 0,
				int32_t x1 = 0, int32_t y1 = 0, int32_t z1 = 0, int32_t w1 = 0,
				int32_t x2 = 0, int32_t y2 = 0, int32_t z2 = 0, int32_t w2 = 0,
				int32_t x3 = 0, int32_t y3 = 0, int32_t z3 = 0, int32_t w3 = 0
			) requires(std::same_as<T, int32_t>&& std::same_as<vector_type, __m512i>) {
				v = _mm512_set_epi32(
					w3, z3, y3, x3,
					w2, z2, y2, x2,
					w1, z1, y1, x1,
					w0, z0, y0, x0
				);
			}

			Vector(
				int64_t x0 = 0, int64_t y0 = 0, int64_t z0 = 0, int64_t w0 = 0,
				int64_t x1 = 0, int64_t y1 = 0, int64_t z1 = 0, int64_t w1 = 0
			) requires(std::same_as<T, int64_t>&& std::same_as<vector_type, __m512i>) {
				v = _mm512_set_epi64(
					w1, z1, y1, x1,
					w0, z0, y0, x0
				);
			}

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<vector_type, __m512h> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_add_ph(reinterpret_cast<__m512>(v), reinterpret_cast<__m512>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<vector_type, __m512h> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_sub_ph(reinterpret_cast<__m512>(v), reinterpret_cast<__m512>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<vector_type, __m512h> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_mul_ph(reinterpret_cast<__m512>(v), reinterpret_cast<__m512>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<vector_type, __m512h> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_div_ph(reinterpret_cast<__m512>(v), reinterpret_cast<__m512>(b.v)));
			}

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<vector_type, __m512> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_add_ps(reinterpret_cast<__m512>(v), reinterpret_cast<__m512>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<vector_type, __m512> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_sub_ps(reinterpret_cast<__m512>(v), reinterpret_cast<__m512>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<vector_type, __m512> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_mul_ps(reinterpret_cast<__m512>(v), reinterpret_cast<__m512>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<vector_type, __m512> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_div_ps(reinterpret_cast<__m512>(v), reinterpret_cast<__m512>(b.v)));
			}

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<vector_type, __m512d> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_add_pd(reinterpret_cast<__m512d>(v), reinterpret_cast<__m512d>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<vector_type, __m512d> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_sub_pd(reinterpret_cast<__m512d>(v), reinterpret_cast<__m512d>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<vector_type, __m512d> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_mul_pd(reinterpret_cast<__m512d>(v), reinterpret_cast<__m512d>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<vector_type, __m512d> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_div_pd(reinterpret_cast<__m512d>(v), reinterpret_cast<__m512d>(b.v)));
			}

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m512i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_add_epi8(reinterpret_cast<__m512i>(v), reinterpret_cast<__m512i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m512i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_sub_epi8(reinterpret_cast<__m512i>(v), reinterpret_cast<__m512i>(b.v)));
			}
			// template <uint32_t blen>
			// Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_mul_epi8(v, bv));
			// }
			// template <uint32_t blen>
			// Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_div_epi8(v, bv));
			// }

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m512i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_add_epi16(reinterpret_cast<__m512i>(v), reinterpret_cast<__m512i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m512i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_sub_epi16(reinterpret_cast<__m512i>(v), reinterpret_cast<__m512i>(b.v)));
			}
			// template <uint32_t blen>
			// Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_mul_epi16(v, bv));
			// }
			// template <uint32_t blen>
			// Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_div_epi16(v, bv));
			// }

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<T, int32_t> && std::same_as<vector_type, __m512i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_add_epi32(reinterpret_cast<__m512i>(v), reinterpret_cast<__m512i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<T, int32_t> && std::same_as<vector_type, __m512i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_sub_epi32(reinterpret_cast<__m512i>(v), reinterpret_cast<__m512i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<T, int32_t> && std::same_as<vector_type, __m512i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_mul_epi32(reinterpret_cast<__m512i>(v), reinterpret_cast<__m512i>(b.v)));
			}
			// template <uint32_t blen>
			// Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<T, i132> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_div_epi32(v, bv));
			// }

			template <uint32_t blen>
			Vector<T, len> operator+(Vector<T, blen> b) requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m512i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_add_epi64(reinterpret_cast<__m512i>(v), reinterpret_cast<__m512i>(b.v)));
			}
			template <uint32_t blen>
			Vector<T, len> operator-(Vector<T, blen> b) requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m512i> && Vector<T, blen>::GetVectorAlignedSize() == GetVectorAlignedSize()) {
				return Vector<T, len>(_mm512_sub_epi64(reinterpret_cast<__m512i>(v), reinterpret_cast<__m512i>(b.v)));
			}
			// template <uint32_t blen>
			// Vector<T, len> operator*(Vector<T, blen> b) requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_mul_epi64(v, bv));
			// }
			// template <uint32_t blen>
			// Vector<T, len> operator/(Vector<T, blen> b) requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m512i>) {
			// 	__m512i v = this->v;
			// 	__m512i bv = b.v;
			// 	return Vector<T, len>(_mm512_div_epi64(v, bv));
			// }

			void Store(T* data) const requires(std::same_as<vector_type, __m512h>) {
				_mm512_storeu_ph(reinterpret_cast<void*>(data), reinterpret_cast<__m512h>(v));
			}
			void Store(T* data) const requires(std::same_as<vector_type, __m512>) {
				_mm512_storeu_ps(reinterpret_cast<void*>(data), reinterpret_cast<__m512>(v));
			}
			void Store(T* data) const requires(std::same_as<vector_type, __m512d>) {
				_mm512_storeu_pd(data, reinterpret_cast<__m512d>(v));
			}
			void Store(T* data) const requires(std::same_as<T, int8_t> && std::same_as<vector_type, __m512i>) {
				_mm512_storeu_epi8(reinterpret_cast<void*>(data), reinterpret_cast<__m512i>(v));
			}
			void Store(T* data) const requires(std::same_as<T, int16_t> && std::same_as<vector_type, __m512i>) {
				_mm512_storeu_epi16(reinterpret_cast<void*>(data), reinterpret_cast<__m512i>(v));
			}
			void Store(T* data) const requires(std::same_as<T, int32_t> && std::same_as<vector_type, __m512i>) {
				_mm512_storeu_epi32(reinterpret_cast<void*>(data), reinterpret_cast<__m512i>(v));
			}
			void Store(T* data) const requires(std::same_as<T, int64_t> && std::same_as<vector_type, __m512i>) {
				_mm512_storeu_epi64(reinterpret_cast<void*>(data), reinterpret_cast<__m512i>(v));
			}
#pragma endregion
	};

	export template <typename T, uint32_t len, uint32_t vectorLenght>
	class VectorVector : public Vector<T, len*vectorLenght> {
		public:
			VectorVector(__m128h v0, __m128h v1) requires(std::same_as<T, __fp16> && vectorLenght*Vector<T, len>::GetVectorAlignedSize() == Vector<T, len*vectorLenght>::GetVectorAlignedSize()) { 
				this->v = _mm256_castps128_ps256(v0);
				this->v = _mm256_insertf128_ps(this->v,v1,1);
			}
			VectorVector(__m128 v0, __m128 v1, __m128 v2, __m128 v3) requires(std::same_as<T, float> && vectorLenght*Vector<T, len>::GetVectorAlignedSize() == Vector<T, len*vectorLenght>::GetVectorAlignedSize()) { 
				this->v = _mm512_castps256_ps512(_mm256_castps128_ps256(v0));
				this->v = _mm512_insertfloatx4(this->v, v1, 1);
				this->v = _mm512_insertfloatx4(this->v, v2, 2);
				this->v = _mm512_insertfloatx4(this->v, v3, 3);
			}
			VectorVector(__m512 v) : Vector<T, len*vectorLenght>(v) { //requires(std::same_as<T, float> && vectorLenght*Vector<T, len>::GetVectorAlignedSize() == Vector<T, len*vectorLenght>::GetVectorAlignedSize()) : Vector<T, len*vectorLenght>(v)  

			}
			VectorVector(
				float x0 = 0, float y0 = 0, float z0 = 0, 
				float x1 = 0, float y1 = 0, float z1 = 0,  
				float x2 = 0, float y2 = 0, float z2 = 0,  
				float x3 = 0, float y3 = 0, float z3 = 0,  
				float x4 = 0, float y4 = 0, float z4 = 0,  
				float x5 = 0
				) requires(std::same_as<T, float> && vectorLenght*Vector<T, len>::GetVectorAlignedSize() == Vector<T, len*vectorLenght>::GetVectorAlignedSize() && len == 3) : 
					Vector<T, len*vectorLenght>(
						x0,y0,z0,
						x1,y1,z1,
						x2,y2,z2,
						x3,y3,z3,
						x4,y4,z4,
						x5) 
				{}

			VectorVector(
				float x0 = 0, float y0 = 0, float z0 = 0, float w0 = 0,
				float x1 = 0, float y1 = 0, float z1 = 0, float w1 = 0,
				float x2 = 0, float y2 = 0, float z2 = 0, float w2 = 0,
				float x3 = 0, float y3 = 0, float z3 = 0, float w3 = 0
			) :
				Vector<T, len*vectorLenght>(
					w3, z3, y3, x3,
					w2, z2, y2, x2,
					w1, z1, y1, x1,
					w0, z0, y0, x0)
			{}
	};

	export Vector<float, 4> g_XMNegIdentityR0(-1.0f, 0.0f, 0.0f, 0.0f);
    export Vector<float, 4> g_XMNegIdentityR1(0.0f, -1.0f, 0.0f, 0.0f);
    export Vector<float, 4> g_XMNegIdentityR2(0.0f, 0.0f, -1.0f, 0.0f);
    export Vector<float, 4> g_XMNegIdentityR3(0.0f, 0.0f, 0.0f, -1.0f);
	export Vector<float, 4> g_XMIdentityR0(1.0f, 0.0f, 0.0f, 0.0f);
    export Vector<float, 4> g_XMIdentityR1(0.0f, 1.0f, 0.0f, 0.0f);
    export Vector<float, 4> g_XMIdentityR2(0.0f, 0.0f, 1.0f, 0.0f);
    export Vector<float, 4> g_XMIdentityR3(0.0f, 0.0f, 0.0f, 1.0f);
	export Vector<uint32_t, 4> g_XMMaskXY(0xFFFFFFFF, 0xFFFFFFFF, 0x00000000, 0x00000000);
    export Vector<uint32_t, 4> g_XMMask3(0xFFFFFFFF, 0xFFFFFFFF, 0xFFFFFFFF, 0x00000000);
    export Vector<uint32_t, 4> g_XMMaskX(0xFFFFFFFF, 0x00000000, 0x00000000, 0x00000000);
    export Vector<uint32_t, 4> g_XMMaskY(0x00000000, 0xFFFFFFFF, 0x00000000, 0x00000000);
    export Vector<uint32_t, 4> g_XMMaskZ(0x00000000, 0x00000000, 0xFFFFFFFF, 0x00000000);
    export Vector<uint32_t, 4> g_XMMaskW( 0x00000000, 0x00000000, 0x00000000, 0xFFFFFFFF);
}