SCAIL2 int8convrot Memory Error

### Custom Node Testing

- [ ] I have tried disabling custom nodes and the issue persists (see [how to disable custom nodes](https://docs.comfy.org/troubleshooting/custom-node-issues#step-1%3A-test-with-all-custom-nodes-disabled) if you need help)

### Your question

fp16 was ok on my machine and still finish the jobs.
half size int8convrot goes OOM.
My comfyui uptodated and LTX int8convrot, Zit, Krea2 are fine.

### Logs

```powershell
SamplerCustom
#19
SamplerCustom

Execution failed

Node threw an error during execution.

# ComfyUI Error Report
## Error Details
- **Node ID:** 19
- **Node Type:** SamplerCustom
- **Exception Type:** torch.AcceleratorError
- **Exception Message:** torch.AcceleratorError: CUDA error: an illegal memory access was encountered
Search for `cudaErrorIllegalAddress' in https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__TYPES.html for more information.
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

## Stack Trace

File "D:\sd\comfyui\ComfyUI_onepack_v0230_Pack4_test\ComfyUI\execution.py", line 547, in execute
comfy.model_management.reset_cast_buffers()
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^

File "D:\sd\comfyui\ComfyUI_onepack_v0230_Pack4_test\ComfyUI\comfy\model_management.py", line 1358, in reset_cast_buffers
offload_stream.synchronize()
~~~~~~~~~~~~~~~~~~~~~~~~~~^^

File "D:\sd\comfyui\ComfyUI_onepack_v0230_Pack4_test\python_embeded\Lib\site-packages\torch\cuda\streams.py", line 108, in synchronize
super().synchronize()
~~~~~~~~~~~~~~~~~~~^^

## System Information
- **ComfyUI Version:** 0.27.0
- **Arguments:** ComfyUI\main.py --windows-standalone-build
- **OS:** win32
- **Python Version:** 3.13.12 (tags/v3.13.12:1cbe481, Feb 3 2026, 18:22:25) [MSC v.1944 64 bit (AMD64)]
- **Embedded Python:** true
- **PyTorch Version:** 2.12.1+cu130
## Devices

- **Name:** cuda:0 NVIDIA GeForce RTX 5090 : cudaMallocAsync
- **Type:** cuda
- **VRAM Total:** 34190458880
- **VRAM Free:** 32437698560
- **Torch VRAM Total:** 67108864
- **Torch VRAM Free:** 33554432

## Logs

Failed to retrieve server logs

## Attached Workflow
Please make sure that workflow does not contain any sensitive information such as API keys or passwords.

Workflow too large. Please manually upload the workflow from local file system.

## Additional Context
(Please add any additional context or steps to reproduce the error here)
```

### Other

_No response_

Start by reproducing the SCAIL2 int8convrot failure through the SamplerCustom node, using the reported ComfyUI, PyTorch, CUDA, and RTX 5090 versions. Trace execution.py around line 547 and comfy/model_management.py around reset_cast_buffers; use CUDA_LAUNCH_BLOCKING=1 to locate the failing operation, and consider the issue done when the workflow no longer triggers the illegal memory access.

Written by the indexing model from the issue text.