diff --git a/kernel/arm64/KERNEL.ARMV8SVE b/kernel/arm64/KERNEL.ARMV8SVE
index 4ff53c6d0..0e51f2c2f 100644
--- a/kernel/arm64/KERNEL.ARMV8SVE
+++ b/kernel/arm64/KERNEL.ARMV8SVE
@@ -84,6 +84,11 @@ DGEMVTKERNEL = gemv_t_sve_v1x3.c
 CGEMVTKERNEL = zgemv_t.S
 ZGEMVTKERNEL = zgemv_t.S
 
+SSYMV_L_KERNEL = symv_L_sve_v1x4.c
+SSYMV_U_KERNEL = symv_U_sve_v1x4.c
+DSYMV_L_KERNEL = symv_L_sve_v1x4.c
+DSYMV_U_KERNEL = symv_U_sve_v1x4.c
+
 SASUMKERNEL    = sasum_thunderx2t99.c
 DASUMKERNEL    = dasum_thunderx2t99.c
 CASUMKERNEL    = casum_thunderx2t99.c
diff --git a/kernel/arm64/KERNEL.NEOVERSEN1 b/kernel/arm64/KERNEL.NEOVERSEN1
index e623814d6..665ebe459 100644
--- a/kernel/arm64/KERNEL.NEOVERSEN1
+++ b/kernel/arm64/KERNEL.NEOVERSEN1
@@ -70,6 +70,10 @@ DGEMVTKERNEL = gemv_t.S
 CGEMVTKERNEL = zgemv_t.S
 ZGEMVTKERNEL = zgemv_t.S
 
+SSYMV_L_KERNEL = symv_L_asimd_4x4.c
+SSYMV_U_KERNEL = symv_U_asimd_4x4.c
+DSYMV_L_KERNEL = symv_L_asimd_4x4.c
+DSYMV_U_KERNEL = symv_U_asimd_4x4.c
 
 SASUMKERNEL    = sasum_thunderx2t99.c
 DASUMKERNEL    = dasum_thunderx2t99.c
diff --git a/kernel/arm64/symv_L_asimd_4x4.c b/kernel/arm64/symv_L_asimd_4x4.c
new file mode 100644
index 000000000..b3d15ba67
--- /dev/null
+++ b/kernel/arm64/symv_L_asimd_4x4.c
@@ -0,0 +1,113 @@
+/***************************************************************************
+Copyright (c) 2025, The OpenBLAS Project
+All rights reserved.
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are
+met:
+
+   1. Redistributions of source code must retain the above copyright
+      notice, this list of conditions and the following disclaimer.
+
+   2. Redistributions in binary form must reproduce the above copyright
+      notice, this list of conditions and the following disclaimer in
+      the documentation and/or other materials provided with the
+      distribution.
+   3. Neither the name of the OpenBLAS project nor the names of
+      its contributors may be used to endorse or promote products
+      derived from this software without specific prior written
+      permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
+USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*****************************************************************************/
+
+#include "symv_microk_asimd_4x4.c"
+
+int CNAME(BLASLONG m, BLASLONG offset, FLOAT alpha, FLOAT *a, BLASLONG lda,
+          FLOAT *x, BLASLONG inc_x, FLOAT *y, BLASLONG inc_y, FLOAT *buffer)
+{
+  BLASLONG i, j;
+  FLOAT temp1, temp2;
+  FLOAT tmp1[4];
+  FLOAT tmp2[4];
+  FLOAT *a0, *a1, *a2, *a3;
+  FLOAT x0, x1, x2, x3;
+  FLOAT *X = x;
+  FLOAT *Y = y;
+
+  if (inc_y != 1) {
+    Y = buffer;
+    COPY_K(m, y, inc_y, Y, 1);
+  }
+  if (inc_x != 1) {
+    if (inc_y != 1) {
+      X = Y + m;
+    } else {
+      X = buffer;
+    }
+    COPY_K(m, x, inc_x, X, 1);
+  }
+
+  BLASLONG offset1 = (offset / 4) * 4;
+  for (j = 0; j < offset1; j+=4) {
+    a0 = &a[j*lda];
+    a1 = a0 + lda;
+    a2 = a1 + lda;
+    a3 = a2 + lda;
+    x0 = X[j];
+    x1 = X[j+1];
+    x2 = X[j+2];
+    x3 = X[j+3];
+    tmp2[0] = a0[j  ]*x0 + a0[j+1]*x1 + a0[j+2]*x2 + a0[j+3]*x3;
+    tmp2[1] = a0[j+1]*x0 + a1[j+1]*x1 + a1[j+2]*x2 + a1[j+3]*x3;
+    tmp2[2] = a0[j+2]*x0 + a1[j+2]*x1 + a2[j+2]*x2 + a2[j+3]*x3;
+    tmp2[3] = a0[j+3]*x0 + a1[j+3]*x1 + a2[j+3]*x2 + a3[j+3]*x3;
+    tmp1[0] = alpha * x0;
+    tmp1[1] = alpha * x1;
+    tmp1[2] = alpha * x2;
+    tmp1[3] = alpha * x3;
+
+    BLASLONG m2 = (m/4)*4;
+    if (m2 > j+4)
+      symv_kernel_4x4(j+4, m2, a0, a1, a2, a3, X, Y, tmp1, tmp2);
+
+    for (i = m2; i < m; i++) {
+      Y[i] += tmp1[0] * a0[i];
+      tmp2[0] += a0[i] * X[i];
+      Y[i] += tmp1[1] * a1[i];
+      tmp2[1] += a1[i] * X[i];
+      Y[i] += tmp1[2] * a2[i];
+      tmp2[2] += a2[i] * X[i];
+      Y[i] += tmp1[3] * a3[i];
+      tmp2[3] += a3[i] * X[i];
+    }
+    Y[j]   += alpha * tmp2[0];
+    Y[j+1] += alpha * tmp2[1];
+    Y[j+2] += alpha * tmp2[2];
+    Y[j+3] += alpha * tmp2[3];
+  }
+
+  for (j = offset1; j < offset; j++) {
+    temp1 = alpha * X[j];
+    temp2 = 0.0;
+    Y[j] += temp1 * a[j*lda+j];
+    for (i = j+1; i < m; i++) {
+      Y[i] += temp1 * a[j*lda+i];
+      temp2 += a[j*lda+i] * X[i];
+    }
+    Y[j] += alpha * temp2;
+  }
+
+  if (inc_y != 1) {
+    COPY_K(m, Y, 1, y, inc_y);
+  }
+  return(0);
+}
diff --git a/kernel/arm64/symv_L_sve_v1x4.c b/kernel/arm64/symv_L_sve_v1x4.c
new file mode 100644
index 000000000..4b9252339
--- /dev/null
+++ b/kernel/arm64/symv_L_sve_v1x4.c
@@ -0,0 +1,103 @@
+/***************************************************************************
+Copyright (c) 2025, The OpenBLAS Project
+All rights reserved.
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are
+met:
+
+   1. Redistributions of source code must retain the above copyright
+      notice, this list of conditions and the following disclaimer.
+
+   2. Redistributions in binary form must reproduce the above copyright
+      notice, this list of conditions and the following disclaimer in
+      the documentation and/or other materials provided with the
+      distribution.
+   3. Neither the name of the OpenBLAS project nor the names of
+      its contributors may be used to endorse or promote products
+      derived from this software without specific prior written
+      permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
+USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*****************************************************************************/
+
+#include "symv_microk_sve_v1x4.c"
+
+int CNAME(BLASLONG m, BLASLONG offset, FLOAT alpha, FLOAT *a, BLASLONG lda,
+          FLOAT *x, BLASLONG inc_x, FLOAT *y, BLASLONG inc_y, FLOAT *buffer)
+{
+  BLASLONG i, j;
+  FLOAT temp1, temp2;
+  FLOAT tmp1[4];
+  FLOAT tmp2[4];
+  FLOAT *a0, *a1, *a2, *a3;
+  FLOAT x0, x1, x2, x3;
+  FLOAT *X = x;
+  FLOAT *Y = y;
+
+  if (inc_y != 1) {
+    Y = buffer;
+    COPY_K(m, y, inc_y, Y, 1);
+  }
+  if (inc_x != 1) {
+    if (inc_y != 1) {
+      X = Y + m;
+    } else {
+      X = buffer;
+    }
+    COPY_K(m, x, inc_x, X, 1);
+  }
+
+  BLASLONG offset1 = (offset / 4) * 4;
+
+  for (j = 0; j < offset1; j+=4) {
+    a0 = &a[j*lda];
+    a1 = a0 + lda;
+    a2 = a1 + lda;
+    a3 = a2 + lda;
+    x0 = X[j];
+    x1 = X[j+1];
+    x2 = X[j+2];
+    x3 = X[j+3];
+    tmp2[0] = a0[j  ]*x0 + a0[j+1]*x1 + a0[j+2]*x2 + a0[j+3]*x3;
+    tmp2[1] = a0[j+1]*x0 + a1[j+1]*x1 + a1[j+2]*x2 + a1[j+3]*x3;
+    tmp2[2] = a0[j+2]*x0 + a1[j+2]*x1 + a2[j+2]*x2 + a2[j+3]*x3;
+    tmp2[3] = a0[j+3]*x0 + a1[j+3]*x1 + a2[j+3]*x2 + a3[j+3]*x3;
+    tmp1[0] = alpha * x0;
+    tmp1[1] = alpha * x1;
+    tmp1[2] = alpha * x2;
+    tmp1[3] = alpha * x3;
+
+    symv_kernel_v1x4(j+4, m, a0, a1, a2, a3, X, Y, tmp1, tmp2);
+
+    Y[j]   += alpha * tmp2[0];
+    Y[j+1] += alpha * tmp2[1];
+    Y[j+2] += alpha * tmp2[2];
+    Y[j+3] += alpha * tmp2[3];
+  }
+
+  for (j = offset1; j < offset; j++) {
+    temp1 = alpha * X[j];
+    temp2 = 0.0;
+    a0 = &a[j*lda];
+    Y[j] += temp1 * a0[j];
+    for (i = j+1; i < m; i++) {
+      Y[i] += temp1 * a0[i];
+      temp2 += a0[i] * X[i];
+    }
+    Y[j] += alpha * temp2;
+  }
+
+  if (inc_y != 1) {
+    COPY_K(m, Y, 1, y, inc_y);
+  }
+  return(0);
+}
diff --git a/kernel/arm64/symv_U_asimd_4x4.c b/kernel/arm64/symv_U_asimd_4x4.c
new file mode 100644
index 000000000..83e954260
--- /dev/null
+++ b/kernel/arm64/symv_U_asimd_4x4.c
@@ -0,0 +1,106 @@
+/***************************************************************************
+Copyright (c) 2025, The OpenBLAS Project
+All rights reserved.
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are
+met:
+
+   1. Redistributions of source code must retain the above copyright
+      notice, this list of conditions and the following disclaimer.
+
+   2. Redistributions in binary form must reproduce the above copyright
+      notice, this list of conditions and the following disclaimer in
+      the documentation and/or other materials provided with the
+      distribution.
+   3. Neither the name of the OpenBLAS project nor the names of
+      its contributors may be used to endorse or promote products
+      derived from this software without specific prior written
+      permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
+USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*****************************************************************************/
+
+#include "symv_microk_asimd_4x4.c"
+
+int CNAME(BLASLONG m, BLASLONG offset, FLOAT alpha, FLOAT *a, BLASLONG lda,
+          FLOAT *x, BLASLONG inc_x, FLOAT *y, BLASLONG inc_y, FLOAT *buffer)
+{
+  BLASLONG i, j, j1, j2, m2;
+  FLOAT temp1, temp2;
+  FLOAT tmp1[4];
+  FLOAT tmp2[4];
+  FLOAT *a0, *a1, *a2, *a3;
+  FLOAT *X = x;
+  FLOAT *Y = y;
+
+  BLASLONG m1 = m - offset;
+  if (inc_y != 1) {
+    Y = buffer;
+    COPY_K(m, y, inc_y, Y, 1);
+  }
+  if (inc_x != 1) {
+    if (inc_y != 1) {
+      X = Y + m;
+    } else {
+      X = buffer;
+    }
+    COPY_K(m, x, inc_x, X, 1);
+  }
+
+  m2 = m - (offset % 4);
+  for (j = m1; j < m2; j += 4) {
+    tmp1[0] = alpha * X[j];
+    tmp1[1] = alpha * X[j+1];
+    tmp1[2] = alpha * X[j+2];
+    tmp1[3] = alpha * X[j+3];
+    tmp2[0] = 0.0;
+    tmp2[1] = 0.0;
+    tmp2[2] = 0.0;
+    tmp2[3] = 0.0;
+    a0 = &a[j*lda];
+    a1 = a0 + lda;
+    a2 = a1 + lda;
+    a3 = a2 + lda;
+    j1 = (j / 4) * 4;
+    if ( j1 )
+      symv_kernel_4x4(0, j1, a0, a1, a2, a3, X, Y, tmp1, tmp2);
+
+    j2 = 0;
+    for (j1 = j ; j1 < j+4 ; j1++) {
+      temp1 = tmp1[j2];
+      temp2 = tmp2[j2];
+      a0 = &a[j1*lda];
+      for (i=j ; i<j1; i++) {
+        Y[i] += temp1 * a0[i];
+        temp2 += a0[i] * X[i];
+      }
+      Y[j1] += temp1 * a0[j1] + alpha * temp2;
+      j2++;
+    }
+  }
+
+  for ( ; j < m; j++) {
+    temp1 = alpha * X[j];
+    temp2 = 0.0;
+    a0 = &a[j*lda];
+    for (i = 0 ; i < j; i++) {
+      Y[i] += temp1 * a0[i];
+      temp2 += a0[i] * X[i];
+    }
+    Y[j] += temp1 * a0[j] + alpha * temp2;
+  }
+
+  if (inc_y != 1) {
+    COPY_K(m, Y, 1, y, inc_y);
+  }
+  return(0);
+}
diff --git a/kernel/arm64/symv_U_sve_v1x4.c b/kernel/arm64/symv_U_sve_v1x4.c
new file mode 100644
index 000000000..b3a151b40
--- /dev/null
+++ b/kernel/arm64/symv_U_sve_v1x4.c
@@ -0,0 +1,104 @@
+/***************************************************************************
+Copyright (c) 2025, The OpenBLAS Project
+All rights reserved.
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are
+met:
+
+   1. Redistributions of source code must retain the above copyright
+      notice, this list of conditions and the following disclaimer.
+
+   2. Redistributions in binary form must reproduce the above copyright
+      notice, this list of conditions and the following disclaimer in
+      the documentation and/or other materials provided with the
+      distribution.
+   3. Neither the name of the OpenBLAS project nor the names of
+      its contributors may be used to endorse or promote products
+      derived from this software without specific prior written
+      permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
+USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*****************************************************************************/
+
+#include "symv_microk_sve_v1x4.c"
+
+int CNAME(BLASLONG m, BLASLONG offset, FLOAT alpha, FLOAT *a, BLASLONG lda,
+          FLOAT *x, BLASLONG inc_x, FLOAT *y, BLASLONG inc_y, FLOAT *buffer)
+{
+  BLASLONG i, j, j1, j2, m2;
+  FLOAT temp1, temp2;
+  FLOAT tmp1[4];
+  FLOAT tmp2[4];
+  FLOAT *a0, *a1, *a2, *a3;
+  FLOAT *X = x;
+  FLOAT *Y = y;
+
+  BLASLONG m1 = m - offset;
+  if (inc_y != 1) {
+    Y = buffer;
+    COPY_K(m, y, inc_y, Y, 1);
+  }
+  if (inc_x != 1) {
+    if (inc_y != 1) {
+      X = Y + m;
+    } else {
+      X = buffer;
+    }
+    COPY_K(m, x, inc_x, X, 1);
+  }
+
+  m2 = m - (offset % 4);
+  for (j = m1; j < m2; j += 4) {
+    tmp1[0] = alpha * X[j];
+    tmp1[1] = alpha * X[j+1];
+    tmp1[2] = alpha * X[j+2];
+    tmp1[3] = alpha * X[j+3];
+    tmp2[0] = 0.0;
+    tmp2[1] = 0.0;
+    tmp2[2] = 0.0;
+    tmp2[3] = 0.0;
+    a0 = &a[j*lda];
+    a1 = a0 + lda;
+    a2 = a1 + lda;
+    a3 = a2 + lda;
+    symv_kernel_v1x4(0, j, a0, a1, a2, a3, X, Y, tmp1, tmp2);
+
+    j2 = 0;
+    for (j1 = j ; j1 < j+4 ; j1++) {
+      temp1 = tmp1[j2];
+      temp2 = tmp2[j2];
+      a0 = &a[j1*lda];
+      for (i=j ; i<j1; i++) {
+        Y[i] += temp1 * a0[i];
+        temp2 += a0[i] * X[i];
+      }
+      Y[j1] += temp1 * a0[j1] + alpha * temp2;
+      j2++;
+    }
+  }
+
+  for ( ; j < m; j++) {
+    temp1 = alpha * X[j];
+    temp2 = 0.0;
+    a0 = &a[j*lda];
+    for (i = 0 ; i < j; i++) {
+      Y[i] += temp1 * a0[i];
+      temp2 += a0[i] * X[i];
+    }
+    Y[j] += temp1 * a0[j] + alpha * temp2;
+  }
+
+  if (inc_y != 1) {
+    COPY_K(m, Y, 1, y, inc_y);
+  }
+  return(0);
+}
diff --git a/kernel/arm64/symv_microk_asimd_4x4.c b/kernel/arm64/symv_microk_asimd_4x4.c
new file mode 100644
index 000000000..16ab12ab4
--- /dev/null
+++ b/kernel/arm64/symv_microk_asimd_4x4.c
@@ -0,0 +1,120 @@
+/***************************************************************************
+Copyright (c) 2025, The OpenBLAS Project
+All rights reserved.
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are
+met:
+
+   1. Redistributions of source code must retain the above copyright
+      notice, this list of conditions and the following disclaimer.
+
+   2. Redistributions in binary form must reproduce the above copyright
+      notice, this list of conditions and the following disclaimer in
+      the documentation and/or other materials provided with the
+      distribution.
+   3. Neither the name of the OpenBLAS project nor the names of
+      its contributors may be used to endorse or promote products
+      derived from this software without specific prior written
+      permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
+USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*****************************************************************************/
+
+#include "common.h"
+#include <arm_neon.h>
+
+static void symv_kernel_4x4(BLASLONG from, BLASLONG to, FLOAT *a0, FLOAT *a1, FLOAT *a2, FLOAT *a3,
+                            FLOAT *x, FLOAT *y, FLOAT *temp1, FLOAT *temp2)
+{
+#ifdef DOUBLE
+  float64x2_t vtmpx0 = vld1q_dup_f64(&temp1[0]);
+  float64x2_t vtmpx1 = vld1q_dup_f64(&temp1[1]);
+  float64x2_t vtmpx2 = vld1q_dup_f64(&temp1[2]);
+  float64x2_t vtmpx3 = vld1q_dup_f64(&temp1[3]);
+  float64x2_t vtmpy0 = {0.0, 0.0};
+  float64x2_t vtmpy1 = {0.0, 0.0};
+  float64x2_t vtmpy2 = {0.0, 0.0};
+  float64x2_t vtmpy3 = {0.0, 0.0};
+  float64x2_t vxl, vxh, vyl, vyh;
+  float64x2_t vap0l, vap0h, vap1l, vap1h, vap2l, vap2h, vap3l, vap3h;
+  BLASLONG i;
+  for (i = from; i < to; i+=4) {
+    vyl = vld1q_f64(&y[i]);
+    vyh = vld1q_f64(&y[i+2]);
+    vxl = vld1q_f64(&x[i]);
+    vxh = vld1q_f64(&x[i+2]);
+    vap0l = vld1q_f64(&a0[i]);
+    vap0h = vld1q_f64(&a0[i+2]);
+    vap1l = vld1q_f64(&a1[i]);
+    vap1h = vld1q_f64(&a1[i+2]);
+    vap2l = vld1q_f64(&a2[i]);
+    vap2h = vld1q_f64(&a2[i+2]);
+    vap3l = vld1q_f64(&a3[i]);
+    vap3h = vld1q_f64(&a3[i+2]);
+    vyl = vfmaq_f64(vyl, vtmpx0, vap0l);
+    vyh = vfmaq_f64(vyh, vtmpx0, vap0h);
+    vyl = vfmaq_f64(vyl, vtmpx1, vap1l);
+    vyh = vfmaq_f64(vyh, vtmpx1, vap1h);
+    vyl = vfmaq_f64(vyl, vtmpx2, vap2l);
+    vyh = vfmaq_f64(vyh, vtmpx2, vap2h);
+    vyl = vfmaq_f64(vyl, vtmpx3, vap3l);
+    vyh = vfmaq_f64(vyh, vtmpx3, vap3h);
+    vtmpy0 = vfmaq_f64(vtmpy0, vxl, vap0l);
+    vtmpy0 = vfmaq_f64(vtmpy0, vxh, vap0h);
+    vtmpy1 = vfmaq_f64(vtmpy1, vxl, vap1l);
+    vtmpy2 = vfmaq_f64(vtmpy2, vxl, vap2l);
+    vtmpy1 = vfmaq_f64(vtmpy1, vxh, vap1h);
+    vtmpy2 = vfmaq_f64(vtmpy2, vxh, vap2h);
+    vtmpy3 = vfmaq_f64(vtmpy3, vxl, vap3l);
+    vtmpy3 = vfmaq_f64(vtmpy3, vxh, vap3h);
+    vst1q_f64(&y[i], vyl);
+    vst1q_f64(&y[i+2], vyh);
+  }
+  temp2[0] += vaddvq_f64(vtmpy0);
+  temp2[1] += vaddvq_f64(vtmpy1);
+  temp2[2] += vaddvq_f64(vtmpy2);
+  temp2[3] += vaddvq_f64(vtmpy3);
+#else
+  float32x4_t vtmpx0 = vld1q_dup_f32(&temp1[0]);
+  float32x4_t vtmpx1 = vld1q_dup_f32(&temp1[1]);
+  float32x4_t vtmpx2 = vld1q_dup_f32(&temp1[2]);
+  float32x4_t vtmpx3 = vld1q_dup_f32(&temp1[3]);
+  float32x4_t vtmpy0 = {0.0, 0.0, 0.0, 0.0};
+  float32x4_t vtmpy1 = {0.0, 0.0, 0.0, 0.0};
+  float32x4_t vtmpy2 = {0.0, 0.0, 0.0, 0.0};
+  float32x4_t vtmpy3 = {0.0, 0.0, 0.0, 0.0};
+  float32x4_t vx, vy;
+  float32x4_t vap0, vap1, vap2, vap3;
+  BLASLONG i;
+  for (i = from; i < to; i+=4) {
+    vy = vld1q_f32(&y[i]);
+    vx = vld1q_f32(&x[i]);
+    vap0 = vld1q_f32(&a0[i]);
+    vap1 = vld1q_f32(&a1[i]);
+    vap2 = vld1q_f32(&a2[i]);
+    vap3 = vld1q_f32(&a3[i]);
+    vy = vfmaq_f32(vy, vtmpx0, vap0);
+    vy = vfmaq_f32(vy, vtmpx1, vap1);
+    vy = vfmaq_f32(vy, vtmpx2, vap2);
+    vy = vfmaq_f32(vy, vtmpx3, vap3);
+    vtmpy0 = vfmaq_f32(vtmpy0, vx, vap0);
+    vtmpy1 = vfmaq_f32(vtmpy1, vx, vap1);
+    vtmpy2 = vfmaq_f32(vtmpy2, vx, vap2);
+    vtmpy3 = vfmaq_f32(vtmpy3, vx, vap3);
+    vst1q_f32(&y[i], vy);
+  }
+  temp2[0] += vaddvq_f32(vtmpy0);
+  temp2[1] += vaddvq_f32(vtmpy1);
+  temp2[2] += vaddvq_f32(vtmpy2);
+  temp2[3] += vaddvq_f32(vtmpy3);
+#endif
+}
diff --git a/kernel/arm64/symv_microk_sve_v1x4.c b/kernel/arm64/symv_microk_sve_v1x4.c
new file mode 100644
index 000000000..f87613f39
--- /dev/null
+++ b/kernel/arm64/symv_microk_sve_v1x4.c
@@ -0,0 +1,89 @@
+/***************************************************************************
+Copyright (c) 2025, The OpenBLAS Project
+All rights reserved.
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are
+met:
+
+   1. Redistributions of source code must retain the above copyright
+      notice, this list of conditions and the following disclaimer.
+
+   2. Redistributions in binary form must reproduce the above copyright
+      notice, this list of conditions and the following disclaimer in
+      the documentation and/or other materials provided with the
+      distribution.
+   3. Neither the name of the OpenBLAS project nor the names of
+      its contributors may be used to endorse or promote products
+      derived from this software without specific prior written
+      permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE
+USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*****************************************************************************/
+
+#include "common.h"
+#include <arm_sve.h>
+
+#ifdef DOUBLE
+#define SV_COUNT svcntd
+#define SV_TYPE svfloat64_t
+#define SV_TRUE svptrue_b64
+#define SV_WHILE svwhilelt_b64_s64
+#define SV_DUP svdup_f64
+#else
+#define SV_COUNT svcntw
+#define SV_TYPE svfloat32_t
+#define SV_TRUE svptrue_b32
+#define SV_WHILE svwhilelt_b32_s64
+#define SV_DUP svdup_f32
+#endif
+
+static void symv_kernel_v1x4(BLASLONG from, BLASLONG to, FLOAT *a0, FLOAT *a1, FLOAT *a2, FLOAT *a3, 
+                             FLOAT *x, FLOAT *y, FLOAT *temp1, FLOAT *temp2)
+{
+  SV_TYPE vtmpx0 = SV_DUP(temp1[0]);
+  SV_TYPE vtmpx1 = SV_DUP(temp1[1]);
+  SV_TYPE vtmpx2 = SV_DUP(temp1[2]);
+  SV_TYPE vtmpx3 = SV_DUP(temp1[3]);
+  SV_TYPE vtmpy0 = SV_DUP(0.0);
+  SV_TYPE vtmpy1 = SV_DUP(0.0);
+  SV_TYPE vtmpy2 = SV_DUP(0.0);
+  SV_TYPE vtmpy3 = SV_DUP(0.0);
+  SV_TYPE vx, vy;
+  SV_TYPE vap0, vap1, vap2, vap3;
+  BLASLONG i;
+  uint64_t sve_size = SV_COUNT();
+  svbool_t pg;
+
+  for (i = from; i < to; i += sve_size) {
+    pg = SV_WHILE(i, to);
+    vy = svld1(pg, &y[i]);
+    vx = svld1(pg, &x[i]);
+    vap0 = svld1(pg, &a0[i]);
+    vap1 = svld1(pg, &a1[i]);
+    vap2 = svld1(pg, &a2[i]);
+    vap3 = svld1(pg, &a3[i]);
+    vy = svmla_m(pg, vy, vtmpx0, vap0);
+    vy = svmla_m(pg, vy, vtmpx1, vap1);
+    vy = svmla_m(pg, vy, vtmpx2, vap2);
+    vy = svmla_m(pg, vy, vtmpx3, vap3);
+    vtmpy0 = svmla_m(pg, vtmpy0, vx, vap0);
+    vtmpy1 = svmla_m(pg, vtmpy1, vx, vap1);
+    vtmpy2 = svmla_m(pg, vtmpy2, vx, vap2);
+    vtmpy3 = svmla_m(pg, vtmpy3, vx, vap3);
+    svst1(pg, &y[i], vy);
+  }
+  pg = SV_TRUE();
+  temp2[0] += svaddv(pg, vtmpy0);
+  temp2[1] += svaddv(pg, vtmpy1);
+  temp2[2] += svaddv(pg, vtmpy2);
+  temp2[3] += svaddv(pg, vtmpy3);
+}