3 ms·
I'm wondering if this approach could mean we'd one day be able to have CUDA running on non-Nvidia GPUs.
by amelius 3y ago
I'm wondering if this approach could mean we'd one day be able to have CUDA running on non-Nvidia GPUs.
- h2odragon 3y ago> Out of scope: Passing through GPUs or subsets of GPU capabilities. Perhaps, but that should probably be a different project. They want a different focus here.