diff --git a/kernel/zarch/caxpy.c b/kernel/zarch/caxpy.c
index e4b484ab7..14a124ae2 100644
--- a/kernel/zarch/caxpy.c
+++ b/kernel/zarch/caxpy.c
@@ -65,6 +65,14 @@ static void caxpy_kernel_16(BLASLONG n, FLOAT *x, FLOAT *y, FLOAT *alpha) {
     "vl   %%v21,80(%%r1,%[y])\n\t"
     "vl   %%v22,96(%%r1,%[y])\n\t"
     "vl   %%v23,112(%%r1,%[y])\n\t"
+    "verllg   %%v24,%%v8,32\n\t"
+    "verllg   %%v25,%%v9,32\n\t"
+    "verllg   %%v26,%%v10,32\n\t"
+    "verllg   %%v27,%%v11,32\n\t"
+    "verllg   %%v28,%%v16,32\n\t"
+    "verllg   %%v29,%%v17,32\n\t"
+    "verllg   %%v30,%%v18,32\n\t"
+    "verllg   %%v31,%%v19,32\n\t"
     "vfmasb %%v8,%%v8,%%v0,%%v12\n\t"
     "vfmasb %%v9,%%v9,%%v0,%%v13\n\t"
     "vfmasb %%v10,%%v10,%%v0,%%v14\n\t"