Minimum hardware requirement to prevent CUDA out of memory
Author: jchai01Created May 25, 2026Updated Jul 6, 2026
Trying to run this on a laptop GPU Model: Dell Precision 3480 GPU: NVIDIA RTX A500 Laptop GPU VRAM: 4GB OS: Ubuntu 22
Getting this error despite trying various flags from the readme, as well as #64
torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 16.00 MiB. GPU 0 has a total capacity of 3.68 GiB of which 27.44 MiB is free. Including non-PyTorch memory, this process has 2.83 GiB memory in use. Of the allocated memory 2.75 GiB is allocated by PyTorch, and 12.77 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)What is the minimum hardware required to get a decent result from the demo? If #64 were to work, is ~21 images/frames a limit I have to work with using a laptop GPU?
Source: Robbyant/lingbot-map