diff --git a/comfy/model_management.py b/comfy/model_management.py
index d880f1970..b1e866c0d 100644
--- a/comfy/model_management.py
+++ b/comfy/model_management.py
@@ -1112,7 +1112,7 @@ def pytorch_attention_enabled():
     return ENABLE_PYTORCH_ATTENTION
 
 def pytorch_attention_enabled_vae():
-    if is_amd():
+    if is_amd() and not amd_min_version(device=None, min_rdna_version=4):
         return False  # enabling pytorch attention on AMD currently causes crash when doing high res
     return pytorch_attention_enabled()