diff --git a/Eigen/src/Core/arch/AVX512/PacketMath.h b/Eigen/src/Core/arch/AVX512/PacketMath.h
index 9e66575a9..9fbb256a1 100644
--- a/Eigen/src/Core/arch/AVX512/PacketMath.h
+++ b/Eigen/src/Core/arch/AVX512/PacketMath.h
@@ -551,7 +551,7 @@ EIGEN_STRONG_INLINE void pstoreu<int>(int* to, const Packet16i& from) {
 template <>
 EIGEN_DEVICE_FUNC inline Packet16f pgather<float, Packet16f>(const float* from,
                                                              Index stride) {
-  Packet16i stride_vector = _mm512_set1_epi32(stride);
+  Packet16i stride_vector = _mm512_set1_epi32(convert_index<int>(stride));
   Packet16i stride_multiplier =
       _mm512_set_epi32(15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
   Packet16i indices = _mm512_mullo_epi32(stride_vector, stride_multiplier);
@@ -561,7 +561,7 @@ EIGEN_DEVICE_FUNC inline Packet16f pgather<float, Packet16f>(const float* from,
 template <>
 EIGEN_DEVICE_FUNC inline Packet8d pgather<double, Packet8d>(const double* from,
                                                             Index stride) {
-  Packet8i stride_vector = _mm256_set1_epi32(stride);
+  Packet8i stride_vector = _mm256_set1_epi32(convert_index<int>(stride));
   Packet8i stride_multiplier = _mm256_set_epi32(7, 6, 5, 4, 3, 2, 1, 0);
   Packet8i indices = _mm256_mullo_epi32(stride_vector, stride_multiplier);
 
@@ -572,7 +572,7 @@ template <>
 EIGEN_DEVICE_FUNC inline void pscatter<float, Packet16f>(float* to,
                                                          const Packet16f& from,
                                                          Index stride) {
-  Packet16i stride_vector = _mm512_set1_epi32(stride);
+  Packet16i stride_vector = _mm512_set1_epi32(convert_index<int>(stride));
   Packet16i stride_multiplier =
       _mm512_set_epi32(15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, 0);
   Packet16i indices = _mm512_mullo_epi32(stride_vector, stride_multiplier);
@@ -582,7 +582,7 @@ template <>
 EIGEN_DEVICE_FUNC inline void pscatter<double, Packet8d>(double* to,
                                                          const Packet8d& from,
                                                          Index stride) {
-  Packet8i stride_vector = _mm256_set1_epi32(stride);
+  Packet8i stride_vector = _mm256_set1_epi32(convert_index<int>(stride));
   Packet8i stride_multiplier = _mm256_set_epi32(7, 6, 5, 4, 3, 2, 1, 0);
   Packet8i indices = _mm256_mullo_epi32(stride_vector, stride_multiplier);
   _mm512_i32scatter_pd(to, indices, from, 8);
@@ -660,8 +660,7 @@ EIGEN_STRONG_INLINE Packet8d pabs(const Packet8d& a) {
 
 #ifdef EIGEN_VECTORIZE_AVX512DQ
 #define EIGEN_INSERT_8f_INTO_16f(OUTPUT, INPUTA, INPUTB) \
-  OUTPUT = _mm512_insertf32x8(OUTPUT, INPUTA, 0);        \
-  OUTPUT = _mm512_insertf32x8(OUTPUT, INPUTB, 1);
+  OUTPUT = _mm512_insertf32x8(_mm512_castps256_ps512(INPUTA), INPUTB, 1);
 #else
 #define EIGEN_INSERT_8f_INTO_16f(OUTPUT, INPUTA, INPUTB)                    \
   OUTPUT = _mm512_insertf32x4(OUTPUT, _mm256_extractf128_ps(INPUTA, 0), 0); \
@@ -855,7 +854,7 @@ template<> EIGEN_STRONG_INLINE Packet8d preduxp<Packet8d>(const Packet8d* vecs)
 
   final_1 = _mm256_add_pd(final_1, _mm256_blend_pd(tmp0, tmp1, 0xC));
 
-  __m512d final_output = _mm512_insertf64x4(final_output, final_0, 0);
+  __m512d final_output = _mm512_castpd256_pd512(final_0);
 
   return _mm512_insertf64x4(final_output, final_1, 1);
 }