[Compile] accelerate compilation speed using NVRTC

[Kathryn-cat]

This PR supports NVRTC as an alternative to NVCC for faster, device-side JIT compilation of CUDA kernels, in favor of the PR apache/tvm-ffi#283.

It enhances the CUDA compilation backend by:

• Adding Python NVRTC support using cuda-python bindings
• Removing legacy C++ NVRTC fallback in favor of a Python-first approach
• Keeping nvcc as the default compiler with fatbin output (no behavior change for existing users)

Users can choose the compilation backend using an environment variable TVM_CUDA_COMPILE_MODE, choosing from "nvcc" and "nvrtc". For example,

TVM_CUDA_COMPILE_MODE=nvrtc python3 your_program.py

Here is a short benchmark of the compilation speed of kernels in test_target_codegen_cuda.py.

NVCC vs NVRTC Compilation Time Comparison (Python-side Call)

Test Case	Code Size	NVCC Time (ms)	NVRTC Time (ms)	Speedup
test_crossthread_reduction1	1945 B	241.27	51.23	4.7x
test_cuda_bf16_vectorize_add	3760 B	342.72	44.50	7.7x
test_cuda_const_float_to_half	12394 B	272.85	31.99	8.5x
test_cuda_device_func_call	975 B	215.58	21.47	10.0x
test_cuda_float_const_hex_format	685 B	217.39	20.52	10.6x
test_cuda_floordiv_with_vectorization	1050 B	213.88	23.32	9.2x
test_cuda_inf_nan	673 B	214.33	24.94	8.6x
test_cuda_tensormap	755 B	213.91	20.74	10.3x
test_cuda_thread_sync_inside_condition	1007 B	213.43	28.29	7.5x
test_cuda_vectorize_add	908 B	226.81	40.39	5.6x
test_cuda_vectorize_load	734 B	217.25	24.02	9.0x
test_device_host_call_same_func	924 B	216.03	21.21	10.2x
test_vectorized_intrin1	847 B	226.15	26.34	8.6x

NVSHMEM Support

Currently, NVSHMEM is not supported via NVRTC.

• Fallback Behavior: When NVSHMEM is required, the compilation pipeline will automatically fall back to NVCC, even if TVM_CUDA_COMPILE_MODE is set to nvrtc.
• Future Roadmap: Support for NVRTC with NVSHMEM is planned for follow-up PRs.

[yzh119]

why not default to nvrtc?

[tqchen]

I think we should cross check the speed diff and once confirmed, we can switch to nvrtc default