diff --git a/python/sglang/srt/disaggregation/ascend/conn.py b/python/sglang/srt/disaggregation/ascend/conn.py
index d323d64d0..2ad9da407 100644
--- a/python/sglang/srt/disaggregation/ascend/conn.py
+++ b/python/sglang/srt/disaggregation/ascend/conn.py
@@ -109,13 +109,17 @@ class AscendKVManager(MooncakeKVManager):
         executor: concurrent.futures.ThreadPoolExecutor,
         dst_layer_ids: Optional[List[int]] = None,
         dst_device_kv_indices: Optional[npt.NDArray[np.int32]] = None,
+        dst_kv_item_len: Optional[int] = None,
+        dst_attn_tp_size: Optional[int] = None,
     ):
         if dst_device_kv_indices is not None:
             raise NotImplementedError(
                 "Ascend PD transfer does not support HiSparse "
                 "destination device KV indices"
             )
-
+        self._validate_envelope_kv_layout(
+            dst_kv_ptrs, dst_kv_item_len, dst_attn_tp_size
+        )
         # Group by indices
         prefill_kv_blocks, dst_kv_blocks = group_concurrent_contiguous(
             prefill_kv_indices, dst_kv_indices
diff --git a/python/sglang/srt/mem_cache/pool_host/mha.py b/python/sglang/srt/mem_cache/pool_host/mha.py
index 2d71a8a08..4646fb255 100644
--- a/python/sglang/srt/mem_cache/pool_host/mha.py
+++ b/python/sglang/srt/mem_cache/pool_host/mha.py
@@ -384,13 +384,18 @@ class MHATokenToKVPoolHost(HostKVCache):
     def backup_from_device_all_layer(
         self, device_pool, host_indices, device_indices, io_backend
     ):
-        (
-            device_k_data_ptrs,
-            device_v_data_ptrs,
-            device_k_buffers,
-            device_v_buffers,
-        ) = self._resolve_device_transfer_buffers(device_pool)
-        device_kv_buffers = device_k_buffers + device_v_buffers
+        if io_backend == "kernel_ascend":
+            # NPU pools use contiguous multi-layer tensors and intentionally do
+            # not build the CUDA-style k_data_ptrs/v_data_ptrs arrays.
+            device_kv_buffers = None
+        else:
+            (
+                device_k_data_ptrs,
+                device_v_data_ptrs,
+                device_k_buffers,
+                device_v_buffers,
+            ) = self._resolve_device_transfer_buffers(device_pool)
+            device_kv_buffers = device_k_buffers + device_v_buffers
         if io_backend == "kernel":
             if self.layout == "layer_first":
                 if self.can_use_jit: