[FP8] Relax the FP8 cuda arch limitation for NVIDIA Ada GPUs (SM89)

[leeeizhang]

Nvidia Ada Lovelace GPUs (e.g., RTX 4090, L20, L40) with SM89 version are also support FP8 MMA, and hence, it is recommended to relax the CUDA architecture limitations to enable FP8 training on a broader range of devices.

and the CUDA 12.0 announcement says that it supports Lovelace architecture:

• CUDA 12.0 exposes programmable functionality for many features of the NVIDIA Hopper and NVIDIA Ada Lovelace architectures: ...32x Ultra xMMA (including FP8 and FP16)

• https://developer.nvidia.com/blog/cuda-toolkit-12-0-released-for-general-availability/
• https://nvidia.github.io/TensorRT-LLM/reference/support-matrix.html
• https://github.com/NVIDIA/cutlass/blob/c4e3e122e266644c61b4af33d0cc09f4c391a64b/include/cutlass/arch/mma_sm89.h#L57