diff --git a/vllm/utils/torch_utils.py b/vllm/utils/torch_utils.py
index 1bff517fd..fe047e0df 100644
--- a/vllm/utils/torch_utils.py
+++ b/vllm/utils/torch_utils.py
@@ -683,7 +683,7 @@ def get_accelerator_view_from_cpu_tensor(cpu_tensor: torch.Tensor) -> torch.Tens
     if current_platform.is_xpu():
         assert cpu_tensor.is_pinned(), "CPU tensor must be pinned"
         return torch.ops._C.get_xpu_view_from_cpu_tensor(cpu_tensor)
-    elif current_platform.is_cuda():
+    elif current_platform.is_cuda() or current_platform.is_rocm():
         return torch.ops._C.get_cuda_view_from_cpu_tensor(cpu_tensor)
     else:
         raise ValueError(