diff --git a/README.md b/README.md
index 52676b5..fc77c1e 100644
--- a/README.md
+++ b/README.md
@@ -14,7 +14,7 @@
 
 | Date | Description |
 | --- | --- |
-| **2026-04-01** | Added node "Voxel to Mesh"<br>It replaces Remeshing to make watertight mesh<br>Added "Blackwell Fix" in "Load Model" node |
+| **2026-04-01** | Added node "Voxel to Mesh"<br>It replaces Remeshing to make watertight mesh |
 | **2026-03-21** | Added node "Projection HighPoly to LowPoly"<br>Added node "Render MultiView" |
 | **2026-03-17** | Added Inpainting Choice NS and TELEA |
 | **2026-03-14** | Added Experimental node "Projection MultiView Texturing"<br>Check in example_workflows folder |
diff --git a/nodes.py b/nodes.py
index 7e05569..7a462c0 100644
--- a/nodes.py
+++ b/nodes.py
@@ -329,7 +329,6 @@ class Trellis2LoadModel:
                 "device": (["cpu","cuda"],{"default":"cuda"}),
                 "low_vram": ("BOOLEAN",{"default":True}),
                 "keep_models_loaded": ("BOOLEAN", {"default":True}),
-                "fix_blackwell": ("BOOLEAN",{"default":False}),
                 "conv_backend": (["spconv","torchsparse","flex_gemm"],{"default":"flex_gemm"}),
                 "sparse_backend": (["xformers","flash_attn"],{"default":"flash_attn"}),
             },
@@ -341,15 +340,7 @@ class Trellis2LoadModel:
     CATEGORY = "Trellis2Wrapper"
     OUTPUT_NODE = True
 
-    def process(self, modelname, backend, device, low_vram, keep_models_loaded, fix_blackwell, conv_backend, sparse_backend):
-        if fix_blackwell:
-            from .blackwell_fix import patch_all
-            patch_all()
-            print('conv_backend is forced to spconv')
-            sparseconfig.set_conv_backend('spconv')
-        else:
-            sparseconfig.set_conv_backend(conv_backend)
-            
+    def process(self, modelname, backend, device, low_vram, keep_models_loaded, conv_backend, sparse_backend):    
         os.environ['OPENCV_IO_ENABLE_OPENEXR'] = '1'
         os.environ["PYTORCH_CUDA_ALLOC_CONF"] = "expandable_segments:True"  # Can save GPU memory
         #os.environ["FLEX_GEMM_AUTOTUNE_CACHE_PATH"] = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'autotune_cache.json')
@@ -358,6 +349,7 @@ class Trellis2LoadModel:
         
         config.set_backend(backend)
         sparseconfig.set_attn_backend(sparse_backend)
+        sparseconfig.set_conv_backend(conv_backend)
         
         reset_cuda()