diff --git a/simde/arm/neon/addl_high.h b/simde/arm/neon/addl_high.h
index fdef796c9..d4fc7ba7e 100644
--- a/simde/arm/neon/addl_high.h
+++ b/simde/arm/neon/addl_high.h
@@ -28,9 +28,8 @@
 #if !defined(SIMDE_ARM_NEON_ADDL_HIGH_H)
 #define SIMDE_ARM_NEON_ADDL_HIGH_H
 
-#include "add.h"
-#include "movl.h"
-#include "movl_high.h"
+#include "addl.h"
+#include "get_high.h"
 #include "types.h"
 
 HEDLEY_DIAGNOSTIC_PUSH
@@ -43,7 +42,7 @@ simde_vaddl_high_s8(simde_int8x16_t a, simde_int8x16_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vaddl_high_s8(a, b);
   #else
-    return simde_vaddq_s16(simde_vmovl_high_s8(a), simde_vmovl_high_s8(b));
+    return simde_vaddl_s8(simde_vget_high_s8(a), simde_vget_high_s8(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -57,7 +56,7 @@ simde_vaddl_high_s16(simde_int16x8_t a, simde_int16x8_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vaddl_high_s16(a, b);
   #else
-    return simde_vaddq_s32(simde_vmovl_high_s16(a), simde_vmovl_high_s16(b));
+    return simde_vaddl_s16(simde_vget_high_s16(a), simde_vget_high_s16(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -71,7 +70,7 @@ simde_vaddl_high_s32(simde_int32x4_t a, simde_int32x4_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vaddl_high_s32(a, b);
   #else
-    return simde_vaddq_s64(simde_vmovl_high_s32(a), simde_vmovl_high_s32(b));
+    return simde_vaddl_s32(simde_vget_high_s32(a), simde_vget_high_s32(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -85,7 +84,7 @@ simde_vaddl_high_u8(simde_uint8x16_t a, simde_uint8x16_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vaddl_high_u8(a, b);
   #else
-    return simde_vaddq_u16(simde_vmovl_high_u8(a), simde_vmovl_high_u8(b));
+    return simde_vaddl_u8(simde_vget_high_u8(a), simde_vget_high_u8(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -99,7 +98,7 @@ simde_vaddl_high_u16(simde_uint16x8_t a, simde_uint16x8_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vaddl_high_u16(a, b);
   #else
-    return simde_vaddq_u32(simde_vmovl_high_u16(a), simde_vmovl_high_u16(b));
+    return simde_vaddl_u16(simde_vget_high_u16(a), simde_vget_high_u16(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -113,7 +112,7 @@ simde_vaddl_high_u32(simde_uint32x4_t a, simde_uint32x4_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vaddl_high_u32(a, b);
   #else
-    return simde_vaddq_u64(simde_vmovl_high_u32(a), simde_vmovl_high_u32(b));
+    return simde_vaddl_u32(simde_vget_high_u32(a), simde_vget_high_u32(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
diff --git a/simde/arm/neon/addw_high.h b/simde/arm/neon/addw_high.h
index 1f2df9052..eb18437e3 100644
--- a/simde/arm/neon/addw_high.h
+++ b/simde/arm/neon/addw_high.h
@@ -28,8 +28,8 @@
 #define SIMDE_ARM_NEON_ADDW_HIGH_H
 
 #include "types.h"
-#include "movl_high.h"
-#include "add.h"
+#include "get_high.h"
+#include "addw.h"
 
 HEDLEY_DIAGNOSTIC_PUSH
 SIMDE_DISABLE_UNWANTED_DIAGNOSTICS
@@ -40,19 +40,8 @@ simde_int16x8_t
 simde_vaddw_high_s8(simde_int16x8_t a, simde_int8x16_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vaddw_high_s8(a, b);
-  #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
-    return simde_vaddq_s16(a, simde_vmovl_high_s8(b));
   #else
-    simde_int16x8_private r_;
-    simde_int16x8_private a_ = simde_int16x8_to_private(a);
-    simde_int8x16_private b_ = simde_int8x16_to_private(b);
-
-    SIMDE_VECTORIZE
-    for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-      r_.values[i] = a_.values[i] + b_.values[i + ((sizeof(b_.values) / sizeof(b_.values[0])) / 2)];
-    }
-
-    return simde_int16x8_from_private(r_);
+    return simde_vaddw_s8(a, simde_vget_high_s8(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -65,19 +54,8 @@ simde_int32x4_t
 simde_vaddw_high_s16(simde_int32x4_t a, simde_int16x8_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vaddw_high_s16(a, b);
-  #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
-    return simde_vaddq_s32(a, simde_vmovl_high_s16(b));
   #else
-    simde_int32x4_private r_;
-    simde_int32x4_private a_ = simde_int32x4_to_private(a);
-    simde_int16x8_private b_ = simde_int16x8_to_private(b);
-
-    SIMDE_VECTORIZE
-    for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-      r_.values[i] = a_.values[i] + b_.values[i + ((sizeof(b_.values) / sizeof(b_.values[0])) / 2)];
-    }
-
-    return simde_int32x4_from_private(r_);
+    return simde_vaddw_s16(a, simde_vget_high_s16(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -90,19 +68,8 @@ simde_int64x2_t
 simde_vaddw_high_s32(simde_int64x2_t a, simde_int32x4_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vaddw_high_s32(a, b);
-  #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
-    return simde_vaddq_s64(a, simde_vmovl_high_s32(b));
   #else
-    simde_int64x2_private r_;
-    simde_int64x2_private a_ = simde_int64x2_to_private(a);
-    simde_int32x4_private b_ = simde_int32x4_to_private(b);
-
-    SIMDE_VECTORIZE
-    for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-      r_.values[i] = a_.values[i] + b_.values[i + ((sizeof(b_.values) / sizeof(b_.values[0])) / 2)];
-    }
-
-    return simde_int64x2_from_private(r_);
+    return simde_vaddw_s32(a, simde_vget_high_s32(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -115,19 +82,8 @@ simde_uint16x8_t
 simde_vaddw_high_u8(simde_uint16x8_t a, simde_uint8x16_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vaddw_high_u8(a, b);
-  #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
-    return simde_vaddq_u16(a, simde_vmovl_high_u8(b));
   #else
-    simde_uint16x8_private r_;
-    simde_uint16x8_private a_ = simde_uint16x8_to_private(a);
-    simde_uint8x16_private b_ = simde_uint8x16_to_private(b);
-
-    SIMDE_VECTORIZE
-    for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-      r_.values[i] = a_.values[i] + b_.values[i + ((sizeof(b_.values) / sizeof(b_.values[0])) / 2)];
-    }
-
-    return simde_uint16x8_from_private(r_);
+    return simde_vaddw_u8(a, simde_vget_high_u8(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -140,19 +96,8 @@ simde_uint32x4_t
 simde_vaddw_high_u16(simde_uint32x4_t a, simde_uint16x8_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vaddw_high_u16(a, b);
-  #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
-    return simde_vaddq_u32(a, simde_vmovl_high_u16(b));
   #else
-    simde_uint32x4_private r_;
-    simde_uint32x4_private a_ = simde_uint32x4_to_private(a);
-    simde_uint16x8_private b_ = simde_uint16x8_to_private(b);
-
-    SIMDE_VECTORIZE
-    for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-      r_.values[i] = a_.values[i] + b_.values[i + ((sizeof(b_.values) / sizeof(b_.values[0])) / 2)];
-    }
-
-    return simde_uint32x4_from_private(r_);
+    return simde_vaddw_u16(a, simde_vget_high_u16(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -165,19 +110,8 @@ simde_uint64x2_t
 simde_vaddw_high_u32(simde_uint64x2_t a, simde_uint32x4_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vaddw_high_u32(a, b);
-  #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
-    return simde_vaddq_u64(a, simde_vmovl_high_u32(b));
   #else
-    simde_uint64x2_private r_;
-    simde_uint64x2_private a_ = simde_uint64x2_to_private(a);
-    simde_uint32x4_private b_ = simde_uint32x4_to_private(b);
-
-    SIMDE_VECTORIZE
-    for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-      r_.values[i] = a_.values[i] + b_.values[i + ((sizeof(b_.values) / sizeof(b_.values[0])) / 2)];
-    }
-
-    return simde_uint64x2_from_private(r_);
+    return simde_vaddw_u32(a, simde_vget_high_u32(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
diff --git a/simde/arm/neon/mlal_high.h b/simde/arm/neon/mlal_high.h
index f7222d16f..9862c740e 100644
--- a/simde/arm/neon/mlal_high.h
+++ b/simde/arm/neon/mlal_high.h
@@ -28,8 +28,8 @@
 #if !defined(SIMDE_ARM_NEON_MLAL_HIGH_H)
 #define SIMDE_ARM_NEON_MLAL_HIGH_H
 
-#include "movl_high.h"
-#include "mla.h"
+#include "get_high.h"
+#include "mlal.h"
 #include "types.h"
 
 HEDLEY_DIAGNOSTIC_PUSH
@@ -42,7 +42,7 @@ simde_vmlal_high_s8(simde_int16x8_t a, simde_int8x16_t b, simde_int8x16_t c) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlal_high_s8(a, b, c);
   #else
-    return simde_vmlaq_s16(a, simde_vmovl_high_s8(b), simde_vmovl_high_s8(c));
+    return simde_vmlal_s8(a, simde_vget_high_s8(b), simde_vget_high_s8(c));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -56,7 +56,7 @@ simde_vmlal_high_s16(simde_int32x4_t a, simde_int16x8_t b, simde_int16x8_t c) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlal_high_s16(a, b, c);
   #else
-    return simde_vmlaq_s32(a, simde_vmovl_high_s16(b), simde_vmovl_high_s16(c));
+    return simde_vmlal_s16(a, simde_vget_high_s16(b), simde_vget_high_s16(c));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -70,22 +70,7 @@ simde_vmlal_high_s32(simde_int64x2_t a, simde_int32x4_t b, simde_int32x4_t c) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlal_high_s32(a, b, c);
   #else
-    simde_int64x2_private
-      r_,
-      a_ = simde_int64x2_to_private(a),
-      b_ = simde_int64x2_to_private(simde_vmovl_high_s32(b)),
-      c_ = simde_int64x2_to_private(simde_vmovl_high_s32(c));
-
-    #if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
-      r_.values = (b_.values * c_.values) + a_.values;
-    #else
-      SIMDE_VECTORIZE
-      for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-        r_.values[i] = (b_.values[i] * c_.values[i]) + a_.values[i];
-      }
-    #endif
-
-    return simde_int64x2_from_private(r_);
+    return simde_vmlal_s32(a, simde_vget_high_s32(b), simde_vget_high_s32(c));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -99,7 +84,7 @@ simde_vmlal_high_u8(simde_uint16x8_t a, simde_uint8x16_t b, simde_uint8x16_t c)
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlal_high_u8(a, b, c);
   #else
-    return simde_vmlaq_u16(a, simde_vmovl_high_u8(b), simde_vmovl_high_u8(c));
+    return simde_vmlal_u8(a, simde_vget_high_u8(b), simde_vget_high_u8(c));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -113,7 +98,7 @@ simde_vmlal_high_u16(simde_uint32x4_t a, simde_uint16x8_t b, simde_uint16x8_t c)
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlal_high_u16(a, b, c);
   #else
-    return simde_vmlaq_u32(a, simde_vmovl_high_u16(b), simde_vmovl_high_u16(c));
+    return simde_vmlal_u16(a, simde_vget_high_u16(b), simde_vget_high_u16(c));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -127,22 +112,7 @@ simde_vmlal_high_u32(simde_uint64x2_t a, simde_uint32x4_t b, simde_uint32x4_t c)
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlal_high_u32(a, b, c);
   #else
-    simde_uint64x2_private
-      r_,
-      a_ = simde_uint64x2_to_private(a),
-      b_ = simde_uint64x2_to_private(simde_vmovl_high_u32(b)),
-      c_ = simde_uint64x2_to_private(simde_vmovl_high_u32(c));
-
-    #if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
-      r_.values = (b_.values * c_.values) + a_.values;
-    #else
-      SIMDE_VECTORIZE
-      for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-        r_.values[i] = (b_.values[i] * c_.values[i]) + a_.values[i];
-      }
-    #endif
-
-    return simde_uint64x2_from_private(r_);
+    return simde_vmlal_u32(a, simde_vget_high_u32(b), simde_vget_high_u32(c));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
diff --git a/simde/arm/neon/mlal_high_n.h b/simde/arm/neon/mlal_high_n.h
index 0c26174ec..915989c1b 100644
--- a/simde/arm/neon/mlal_high_n.h
+++ b/simde/arm/neon/mlal_high_n.h
@@ -27,9 +27,8 @@
 #if !defined(SIMDE_ARM_NEON_MLAL_HIGH_N_H)
 #define SIMDE_ARM_NEON_MLAL_HIGH_N_H
 
-#include "movl_high.h"
-#include "dup_n.h"
-#include "mla.h"
+#include "get_high.h"
+#include "mlal_n.h"
 #include "types.h"
 
 HEDLEY_DIAGNOSTIC_PUSH
@@ -42,7 +41,7 @@ simde_vmlal_high_n_s16(simde_int32x4_t a, simde_int16x8_t b, int16_t c) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlal_high_n_s16(a, b, c);
   #else
-    return simde_vmlaq_s32(a, simde_vmovl_high_s16(b), simde_vdupq_n_s32(c));
+    return simde_vmlal_n_s16(a, simde_vget_high_s16(b), c);
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -56,22 +55,7 @@ simde_vmlal_high_n_s32(simde_int64x2_t a, simde_int32x4_t b, int32_t c) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlal_high_n_s32(a, b, c);
   #else
-    simde_int64x2_private
-      r_,
-      a_ = simde_int64x2_to_private(a),
-      b_ = simde_int64x2_to_private(simde_vmovl_high_s32(b)),
-      c_ = simde_int64x2_to_private(simde_vdupq_n_s64(c));
-
-    #if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
-      r_.values = (b_.values * c_.values) + a_.values;
-    #else
-      SIMDE_VECTORIZE
-      for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-        r_.values[i] = (b_.values[i] * c_.values[i]) + a_.values[i];
-      }
-    #endif
-
-    return simde_int64x2_from_private(r_);
+    return simde_vmlal_n_s32(a, simde_vget_high_s32(b), c);
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -85,7 +69,7 @@ simde_vmlal_high_n_u16(simde_uint32x4_t a, simde_uint16x8_t b, uint16_t c) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlal_high_n_u16(a, b, c);
   #else
-    return simde_vmlaq_u32(a, simde_vmovl_high_u16(b), simde_vdupq_n_u32(c));
+    return simde_vmlal_n_u16(a, simde_vget_high_u16(b), c);
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -99,22 +83,7 @@ simde_vmlal_high_n_u32(simde_uint64x2_t a, simde_uint32x4_t b, uint32_t c) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlal_high_n_u32(a, b, c);
   #else
-    simde_uint64x2_private
-      r_,
-      a_ = simde_uint64x2_to_private(a),
-      b_ = simde_uint64x2_to_private(simde_vmovl_high_u32(b)),
-      c_ = simde_uint64x2_to_private(simde_vdupq_n_u64(c));
-
-    #if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
-      r_.values = (b_.values * c_.values) + a_.values;
-    #else
-      SIMDE_VECTORIZE
-      for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-        r_.values[i] = (b_.values[i] * c_.values[i]) + a_.values[i];
-      }
-    #endif
-
-    return simde_uint64x2_from_private(r_);
+    return simde_vmlal_n_u32(a, simde_vget_high_u32(b), c);
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
diff --git a/simde/arm/neon/mlsl_high.h b/simde/arm/neon/mlsl_high.h
index d70ca935d..d20cb4274 100644
--- a/simde/arm/neon/mlsl_high.h
+++ b/simde/arm/neon/mlsl_high.h
@@ -27,8 +27,8 @@
 #if !defined(SIMDE_ARM_NEON_MLSL_HIGH_H)
 #define SIMDE_ARM_NEON_MLSL_HIGH_H
 
-#include "mull_high.h"
-#include "sub.h"
+#include "get_high.h"
+#include "mlsl.h"
 
 HEDLEY_DIAGNOSTIC_PUSH
 SIMDE_DISABLE_UNWANTED_DIAGNOSTICS
@@ -40,7 +40,7 @@ simde_vmlsl_high_s8(simde_int16x8_t a, simde_int8x16_t b, simde_int8x16_t c) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlsl_high_s8(a, b, c);
   #else
-    return simde_vsubq_s16(a, simde_vmull_high_s8(b, c));
+    return simde_vmlsl_s8(a, simde_vget_high_s8(b), simde_vget_high_s8(c));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -54,7 +54,7 @@ simde_vmlsl_high_s16(simde_int32x4_t a, simde_int16x8_t b, simde_int16x8_t c) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlsl_high_s16(a, b, c);
   #else
-    return simde_vsubq_s32(a, simde_vmull_high_s16(b, c));
+    return simde_vmlsl_s16(a, simde_vget_high_s16(b), simde_vget_high_s16(c));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -68,7 +68,7 @@ simde_vmlsl_high_s32(simde_int64x2_t a, simde_int32x4_t b, simde_int32x4_t c) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlsl_high_s32(a, b, c);
   #else
-    return simde_vsubq_s64(a, simde_vmull_high_s32(b, c));
+    return simde_vmlsl_s32(a, simde_vget_high_s32(b), simde_vget_high_s32(c));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -82,7 +82,7 @@ simde_vmlsl_high_u8(simde_uint16x8_t a, simde_uint8x16_t b, simde_uint8x16_t c)
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlsl_high_u8(a, b, c);
   #else
-    return simde_vsubq_u16(a, simde_vmull_high_u8(b, c));
+    return simde_vmlsl_u8(a, simde_vget_high_u8(b), simde_vget_high_u8(c));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -96,7 +96,7 @@ simde_vmlsl_high_u16(simde_uint32x4_t a, simde_uint16x8_t b, simde_uint16x8_t c)
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlsl_high_u16(a, b, c);
   #else
-    return simde_vsubq_u32(a, simde_vmull_high_u16(b, c));
+    return simde_vmlsl_u16(a, simde_vget_high_u16(b), simde_vget_high_u16(c));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -110,7 +110,7 @@ simde_vmlsl_high_u32(simde_uint64x2_t a, simde_uint32x4_t b, simde_uint32x4_t c)
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlsl_high_u32(a, b, c);
   #else
-    return simde_vsubq_u64(a, simde_vmull_high_u32(b, c));
+    return simde_vmlsl_u32(a, simde_vget_high_u32(b), simde_vget_high_u32(c));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
diff --git a/simde/arm/neon/mlsl_high_n.h b/simde/arm/neon/mlsl_high_n.h
index 7be34c81b..9a1ad1c37 100644
--- a/simde/arm/neon/mlsl_high_n.h
+++ b/simde/arm/neon/mlsl_high_n.h
@@ -27,9 +27,8 @@
 #if !defined(SIMDE_ARM_NEON_MLSL_HIGH_N_H)
 #define SIMDE_ARM_NEON_MLSL_HIGH_N_H
 
-#include "movl_high.h"
-#include "dup_n.h"
-#include "mls.h"
+#include "get_high.h"
+#include "mlsl_n.h"
 #include "types.h"
 
 HEDLEY_DIAGNOSTIC_PUSH
@@ -42,7 +41,7 @@ simde_vmlsl_high_n_s16(simde_int32x4_t a, simde_int16x8_t b, int16_t c) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlsl_high_n_s16(a, b, c);
   #else
-    return simde_vmlsq_s32(a, simde_vmovl_high_s16(b), simde_vdupq_n_s32(c));
+    return simde_vmlsl_n_s16(a, simde_vget_high_s16(b), c);
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -56,22 +55,7 @@ simde_vmlsl_high_n_s32(simde_int64x2_t a, simde_int32x4_t b, int32_t c) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlsl_high_n_s32(a, b, c);
   #else
-    simde_int64x2_private
-      r_,
-      a_ = simde_int64x2_to_private(a),
-      b_ = simde_int64x2_to_private(simde_vmovl_high_s32(b)),
-      c_ = simde_int64x2_to_private(simde_vdupq_n_s64(c));
-
-    #if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
-      r_.values = a_.values - (b_.values * c_.values);
-    #else
-      SIMDE_VECTORIZE
-      for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-        r_.values[i] = a_.values[i] - (b_.values[i] * c_.values[i]);
-      }
-    #endif
-
-    return simde_int64x2_from_private(r_);
+    return simde_vmlsl_n_s32(a, simde_vget_high_s32(b), c);
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -85,7 +69,7 @@ simde_vmlsl_high_n_u16(simde_uint32x4_t a, simde_uint16x8_t b, uint16_t c) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlsl_high_n_u16(a, b, c);
   #else
-    return simde_vmlsq_u32(a, simde_vmovl_high_u16(b), simde_vdupq_n_u32(c));
+    return simde_vmlsl_n_u16(a, simde_vget_high_u16(b), c);
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -99,22 +83,7 @@ simde_vmlsl_high_n_u32(simde_uint64x2_t a, simde_uint32x4_t b, uint32_t c) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmlsl_high_n_u32(a, b, c);
   #else
-    simde_uint64x2_private
-      r_,
-      a_ = simde_uint64x2_to_private(a),
-      b_ = simde_uint64x2_to_private(simde_vmovl_high_u32(b)),
-      c_ = simde_uint64x2_to_private(simde_vdupq_n_u64(c));
-
-    #if defined(SIMDE_VECTOR_SUBSCRIPT_OPS)
-      r_.values = a_.values - (b_.values * c_.values);
-    #else
-      SIMDE_VECTORIZE
-      for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-        r_.values[i] = a_.values[i] - (b_.values[i] * c_.values[i]);
-      }
-    #endif
-
-    return simde_uint64x2_from_private(r_);
+    return simde_vmlsl_n_u32(a, simde_vget_high_u32(b), c);
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
diff --git a/simde/arm/neon/mull_high.h b/simde/arm/neon/mull_high.h
index 87e83369a..6f262e7f4 100644
--- a/simde/arm/neon/mull_high.h
+++ b/simde/arm/neon/mull_high.h
@@ -29,9 +29,9 @@
 #define SIMDE_ARM_NEON_MULL_HIGH_H
 
 #include "types.h"
-#include "mul.h"
-#include "movl_high.h"
 #include "mull.h"
+#include "get_high.h"
+
 
 HEDLEY_DIAGNOSTIC_PUSH
 SIMDE_DISABLE_UNWANTED_DIAGNOSTICS
@@ -43,7 +43,7 @@ simde_vmull_high_s8(simde_int8x16_t a, simde_int8x16_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmull_high_s8(a, b);
   #else
-    return simde_vmulq_s16(simde_vmovl_high_s8(a), simde_vmovl_high_s8(b));
+    return simde_vmull_s8(simde_vget_high_s8(a), simde_vget_high_s8(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -57,7 +57,7 @@ simde_vmull_high_s16(simde_int16x8_t a, simde_int16x8_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmull_high_s16(a, b);
   #else
-    return simde_vmulq_s32(simde_vmovl_high_s16(a), simde_vmovl_high_s16(b));
+    return simde_vmull_s16(simde_vget_high_s16(a), simde_vget_high_s16(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -71,7 +71,7 @@ simde_vmull_high_s32(simde_int32x4_t a, simde_int32x4_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmull_high_s32(a, b);
   #else
-    return simde_x_vmulq_s64(simde_vmovl_high_s32(a), simde_vmovl_high_s32(b));
+    return simde_vmull_s32(simde_vget_high_s32(a), simde_vget_high_s32(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -85,7 +85,7 @@ simde_vmull_high_u8(simde_uint8x16_t a, simde_uint8x16_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmull_high_u8(a, b);
   #else
-    return simde_vmulq_u16(simde_vmovl_high_u8(a), simde_vmovl_high_u8(b));
+    return simde_vmull_u8(simde_vget_high_u8(a), simde_vget_high_u8(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -99,7 +99,7 @@ simde_vmull_high_u16(simde_uint16x8_t a, simde_uint16x8_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmull_high_u16(a, b);
   #else
-    return simde_vmulq_u32(simde_vmovl_high_u16(a), simde_vmovl_high_u16(b));
+    return simde_vmull_u16(simde_vget_high_u16(a), simde_vget_high_u16(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -113,7 +113,7 @@ simde_vmull_high_u32(simde_uint32x4_t a, simde_uint32x4_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vmull_high_u32(a, b);
   #else
-    return simde_x_vmulq_u64(simde_vmovl_high_u32(a), simde_vmovl_high_u32(b));
+    return simde_vmull_u32(simde_vget_high_u32(a), simde_vget_high_u32(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
diff --git a/simde/arm/neon/subl_high.h b/simde/arm/neon/subl_high.h
index d45f4989b..8d003345a 100644
--- a/simde/arm/neon/subl_high.h
+++ b/simde/arm/neon/subl_high.h
@@ -27,9 +27,8 @@
 #if !defined(SIMDE_ARM_NEON_SUBL_HIGH_H)
 #define SIMDE_ARM_NEON_SUBL_HIGH_H
 
-#include "sub.h"
-#include "movl.h"
-#include "movl_high.h"
+#include "subl.h"
+#include "get_high.h"
 #include "types.h"
 
 HEDLEY_DIAGNOSTIC_PUSH
@@ -42,7 +41,7 @@ simde_vsubl_high_s8(simde_int8x16_t a, simde_int8x16_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vsubl_high_s8(a, b);
   #else
-    return simde_vsubq_s16(simde_vmovl_high_s8(a), simde_vmovl_high_s8(b));
+    return simde_vsubl_s8(simde_vget_high_s8(a), simde_vget_high_s8(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -56,7 +55,7 @@ simde_vsubl_high_s16(simde_int16x8_t a, simde_int16x8_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vsubl_high_s16(a, b);
   #else
-    return simde_vsubq_s32(simde_vmovl_high_s16(a), simde_vmovl_high_s16(b));
+    return simde_vsubl_s16(simde_vget_high_s16(a), simde_vget_high_s16(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -70,7 +69,7 @@ simde_vsubl_high_s32(simde_int32x4_t a, simde_int32x4_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vsubl_high_s32(a, b);
   #else
-    return simde_vsubq_s64(simde_vmovl_high_s32(a), simde_vmovl_high_s32(b));
+    return simde_vsubl_s32(simde_vget_high_s32(a), simde_vget_high_s32(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -84,7 +83,7 @@ simde_vsubl_high_u8(simde_uint8x16_t a, simde_uint8x16_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vsubl_high_u8(a, b);
   #else
-    return simde_vsubq_u16(simde_vmovl_high_u8(a), simde_vmovl_high_u8(b));
+    return simde_vsubl_u8(simde_vget_high_u8(a), simde_vget_high_u8(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -98,7 +97,7 @@ simde_vsubl_high_u16(simde_uint16x8_t a, simde_uint16x8_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vsubl_high_u16(a, b);
   #else
-    return simde_vsubq_u32(simde_vmovl_high_u16(a), simde_vmovl_high_u16(b));
+    return simde_vsubl_u16(simde_vget_high_u16(a), simde_vget_high_u16(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -112,7 +111,7 @@ simde_vsubl_high_u32(simde_uint32x4_t a, simde_uint32x4_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vsubl_high_u32(a, b);
   #else
-    return simde_vsubq_u64(simde_vmovl_high_u32(a), simde_vmovl_high_u32(b));
+    return simde_vsubl_u32(simde_vget_high_u32(a), simde_vget_high_u32(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
diff --git a/simde/arm/neon/subw_high.h b/simde/arm/neon/subw_high.h
index 729a478a7..e09551f44 100644
--- a/simde/arm/neon/subw_high.h
+++ b/simde/arm/neon/subw_high.h
@@ -28,8 +28,8 @@
 #define SIMDE_ARM_NEON_SUBW_HIGH_H
 
 #include "types.h"
-#include "movl_high.h"
-#include "sub.h"
+#include "get_high.h"
+#include "subw.h"
 
 HEDLEY_DIAGNOSTIC_PUSH
 SIMDE_DISABLE_UNWANTED_DIAGNOSTICS
@@ -40,24 +40,8 @@ simde_int16x8_t
 simde_vsubw_high_s8(simde_int16x8_t a, simde_int8x16_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vsubw_high_s8(a, b);
-  #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
-    return simde_vsubq_s16(a, simde_vmovl_high_s8(b));
   #else
-    simde_int16x8_private r_;
-    simde_int16x8_private a_ = simde_int16x8_to_private(a);
-    simde_int8x16_private b_ = simde_int8x16_to_private(b);
-
-    #if (SIMDE_NATURAL_VECTOR_SIZE > 0) && defined(SIMDE_VECTOR_SUBSCRIPT_OPS) && defined(SIMDE_CONVERT_VECTOR_)
-      SIMDE_CONVERT_VECTOR_(r_.values, b_.values);
-      r_.values -= a_.values;
-    #else
-      SIMDE_VECTORIZE
-      for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-        r_.values[i] = a_.values[i] - b_.values[i + ((sizeof(b_.values) / sizeof(b_.values[0])) / 2)];
-      }
-    #endif
-
-    return simde_int16x8_from_private(r_);
+    return simde_vsubw_s8(a, simde_vget_high_s8(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -70,24 +54,8 @@ simde_int32x4_t
 simde_vsubw_high_s16(simde_int32x4_t a, simde_int16x8_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vsubw_high_s16(a, b);
-  #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
-    return simde_vsubq_s32(a, simde_vmovl_high_s16(b));
   #else
-    simde_int32x4_private r_;
-    simde_int32x4_private a_ = simde_int32x4_to_private(a);
-    simde_int16x8_private b_ = simde_int16x8_to_private(b);
-
-    #if (SIMDE_NATURAL_VECTOR_SIZE > 0) && defined(SIMDE_VECTOR_SUBSCRIPT_OPS) && defined(SIMDE_CONVERT_VECTOR_)
-      SIMDE_CONVERT_VECTOR_(r_.values, b_.values);
-      r_.values -= a_.values;
-    #else
-      SIMDE_VECTORIZE
-      for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-        r_.values[i] = a_.values[i] - b_.values[i + ((sizeof(b_.values) / sizeof(b_.values[0])) / 2)];
-      }
-    #endif
-
-    return simde_int32x4_from_private(r_);
+    return simde_vsubw_s16(a, simde_vget_high_s16(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -100,24 +68,8 @@ simde_int64x2_t
 simde_vsubw_high_s32(simde_int64x2_t a, simde_int32x4_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vsubw_high_s32(a, b);
-  #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
-    return simde_vsubq_s64(a, simde_vmovl_high_s32(b));
   #else
-    simde_int64x2_private r_;
-    simde_int64x2_private a_ = simde_int64x2_to_private(a);
-    simde_int32x4_private b_ = simde_int32x4_to_private(b);
-
-    #if (SIMDE_NATURAL_VECTOR_SIZE > 0) && defined(SIMDE_VECTOR_SUBSCRIPT_OPS) && defined(SIMDE_CONVERT_VECTOR_)
-      SIMDE_CONVERT_VECTOR_(r_.values, b_.values);
-      r_.values -= a_.values;
-    #else
-      SIMDE_VECTORIZE
-      for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-        r_.values[i] = a_.values[i] - b_.values[i + ((sizeof(b_.values) / sizeof(b_.values[0])) / 2)];
-      }
-    #endif
-
-    return simde_int64x2_from_private(r_);
+    return simde_vsubw_s32(a, simde_vget_high_s32(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -130,24 +82,8 @@ simde_uint16x8_t
 simde_vsubw_high_u8(simde_uint16x8_t a, simde_uint8x16_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vsubw_high_u8(a, b);
-  #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
-    return simde_vsubq_u16(a, simde_vmovl_high_u8(b));
   #else
-    simde_uint16x8_private r_;
-    simde_uint16x8_private a_ = simde_uint16x8_to_private(a);
-    simde_uint8x16_private b_ = simde_uint8x16_to_private(b);
-
-    #if (SIMDE_NATURAL_VECTOR_SIZE > 0) && defined(SIMDE_VECTOR_SUBSCRIPT_OPS) && defined(SIMDE_CONVERT_VECTOR_)
-      SIMDE_CONVERT_VECTOR_(r_.values, b_.values);
-      r_.values -= a_.values;
-    #else
-      SIMDE_VECTORIZE
-      for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-        r_.values[i] = a_.values[i] - b_.values[i + ((sizeof(b_.values) / sizeof(b_.values[0])) / 2)];
-      }
-    #endif
-
-    return simde_uint16x8_from_private(r_);
+    return simde_vsubw_u8(a, simde_vget_high_u8(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -160,24 +96,8 @@ simde_uint32x4_t
 simde_vsubw_high_u16(simde_uint32x4_t a, simde_uint16x8_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vsubw_high_u16(a, b);
-  #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
-    return simde_vsubq_u32(a, simde_vmovl_high_u16(b));
   #else
-    simde_uint32x4_private r_;
-    simde_uint32x4_private a_ = simde_uint32x4_to_private(a);
-    simde_uint16x8_private b_ = simde_uint16x8_to_private(b);
-
-    #if (SIMDE_NATURAL_VECTOR_SIZE > 0) && defined(SIMDE_VECTOR_SUBSCRIPT_OPS) && defined(SIMDE_CONVERT_VECTOR_)
-      SIMDE_CONVERT_VECTOR_(r_.values, b_.values);
-      r_.values -= a_.values;
-    #else
-      SIMDE_VECTORIZE
-      for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-        r_.values[i] = a_.values[i] - b_.values[i + ((sizeof(b_.values) / sizeof(b_.values[0])) / 2)];
-      }
-    #endif
-
-    return simde_uint32x4_from_private(r_);
+    return simde_vsubw_u16(a, simde_vget_high_u16(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)
@@ -190,24 +110,8 @@ simde_uint64x2_t
 simde_vsubw_high_u32(simde_uint64x2_t a, simde_uint32x4_t b) {
   #if defined(SIMDE_ARM_NEON_A64V8_NATIVE)
     return vsubw_high_u32(a, b);
-  #elif SIMDE_NATURAL_VECTOR_SIZE_GE(128)
-    return simde_vsubq_u64(a, simde_vmovl_high_u32(b));
   #else
-    simde_uint64x2_private r_;
-    simde_uint64x2_private a_ = simde_uint64x2_to_private(a);
-    simde_uint32x4_private b_ = simde_uint32x4_to_private(b);
-
-    #if (SIMDE_NATURAL_VECTOR_SIZE > 0) && defined(SIMDE_VECTOR_SUBSCRIPT_OPS) && defined(SIMDE_CONVERT_VECTOR_)
-      SIMDE_CONVERT_VECTOR_(r_.values, b_.values);
-      r_.values -= a_.values;
-    #else
-      SIMDE_VECTORIZE
-      for (size_t i = 0 ; i < (sizeof(r_.values) / sizeof(r_.values[0])) ; i++) {
-        r_.values[i] = a_.values[i] - b_.values[i + ((sizeof(b_.values) / sizeof(b_.values[0])) / 2)];
-      }
-    #endif
-
-    return simde_uint64x2_from_private(r_);
+    return simde_vsubw_u32(a, simde_vget_high_u32(b));
   #endif
 }
 #if defined(SIMDE_ARM_NEON_A64V8_ENABLE_NATIVE_ALIASES)