diff --git a/csrc/cpu/cpu_attn.cpp b/csrc/cpu/cpu_attn.cpp
index 50f17c758c..c92871b076 100644
--- a/csrc/cpu/cpu_attn.cpp
+++ b/csrc/cpu/cpu_attn.cpp
@@ -24,7 +24,9 @@
     switch (HEAD_DIM) {                                         \
       CPU_ATTN_DISPATCH_CASE(32, __VA_ARGS__)                   \
       CPU_ATTN_DISPATCH_CASE(64, __VA_ARGS__)                   \
+      CPU_ATTN_DISPATCH_CASE(80, __VA_ARGS__)	                \
       CPU_ATTN_DISPATCH_CASE(96, __VA_ARGS__)                   \
+      CPU_ATTN_DISPATCH_CASE(112, __VA_ARGS__)                  \
       CPU_ATTN_DISPATCH_CASE(128, __VA_ARGS__)                  \
       CPU_ATTN_DISPATCH_CASE(160, __VA_ARGS__)                  \
       CPU_ATTN_DISPATCH_CASE(192, __VA_ARGS__)                  \
diff --git a/csrc/cpu/cpu_attn_amx.hpp b/csrc/cpu/cpu_attn_amx.hpp
index 8da458b991..129f6423dd 100644
--- a/csrc/cpu/cpu_attn_amx.hpp
+++ b/csrc/cpu/cpu_attn_amx.hpp
@@ -377,7 +377,7 @@ class AttentionImpl<ISA::AMX, scalar_t, head_dim> {
       const int32_t q_heads_per_kv, const int64_t q_num_stride,
       const int64_t q_head_stride, const float scale) {
     constexpr int64_t bytes_per_head = head_dim * sizeof(scalar_t);
-    static_assert(bytes_per_head % AMX_TILE_ROW_BYTES == 0);
+ // static_assert(bytes_per_head % AMX_TILE_ROW_BYTES == 0);
     constexpr int64_t head_size_block_num = bytes_per_head / AMX_TILE_ROW_BYTES;
     constexpr int64_t head_elem_num_pre_block =
         AMX_TILE_ROW_BYTES / sizeof(scalar_t);
diff --git a/vllm/v1/attention/backends/cpu_attn.py b/vllm/v1/attention/backends/cpu_attn.py
index f1254352c0..eb087fc9ee 100644
--- a/vllm/v1/attention/backends/cpu_attn.py
+++ b/vllm/v1/attention/backends/cpu_attn.py
@@ -42,7 +42,7 @@ def get_supported_dtypes(cls) -> list[torch.dtype]:
 
     @classmethod
     def get_supported_head_sizes(cls) -> list[int]:
-        return [32, 64, 96, 128, 160, 192, 224, 256]
+        return [32, 64, 80, 96, 112, 128, 160, 192, 224, 256]
 
     @staticmethod
     def get_name() -> str:
@@ -137,7 +137,7 @@ def __init__(
         if self.window_size is None:
             self.window_size = -1
         self.block_size = vllm_config.cache_config.block_size
-        self.isa = _get_attn_isa(self.dtype, self.block_size)
+        self.isa = _get_attn_isa(self.dtype, self.block_size, self.head_dim)
 
     def build(
         self,
@@ -486,7 +486,9 @@ def _make_sliding_window_bias(
     return attn_biases
 
 
-def _get_attn_isa(dtype: torch.dtype, block_size: int) -> str:
+def _get_attn_isa(dtype: torch.dtype, block_size: int, head_size: int | None = None) -> str:
+    if head_size is not None and head_size % 32 != 0 and head_size % 16 == 0:
+        return "vec16"
     supports_amx = torch._C._cpu._is_amx_tile_supported()
     if supports_amx and dtype in (torch.bfloat16,) and block_size % 32 == 0:
         return "amx"