3 ms·
https://arxiv.org/abs/2205.14135 https://arxiv.org/abs/2205.14135 - Section 5 suggests that the biggest limitation is that custom CUDA kernels need to be coded
by lhl 4y ago
https://arxiv.org/abs/2205.14135 https://arxiv.org/abs/2205.14135 - Section 5 suggests that the biggest limitation is that custom CUDA kernels need to be coded on a per-GPU architecture basis.