Merge multi-platform hardware benchmarks

[baominghelly]

Summary

• merge the hardware_adapt and cambricon_test histories
• preserve the existing CUDA adapter API, CLI shape, parsers, and metric names
• add platform selection, builds, and parsing for MetaX, Iluvatar CoreX, Hygon DCU, Moore Threads, Cambricon, and Ascend
• fix native device selection, Cambricon cache execution/CNRT compatibility, Ascend hierarchy metadata, MetaX packaged-toolchain builds, and Moore Threads inventory parsing
• document multi-platform build, device visibility, and container requirements

Validation

• 34 passed on A100
• Python compile checks, shell syntax checks, and git diff --check pass
• CUDA 13/A100: full memory, STREAM, L1, and L2 suite passed; final-commit STREAM passed
• MetaX C550, Iluvatar TG-V200, Hygon Z100L, Moore Threads S5000, and Cambricon MLU590: build plus short STREAM passed
• Ascend 910B4: build and short STREAM completed successfully; the development container returned its known exit code 137 after printing completion
• Moore Threads inventory confirmed on hardware as 8 x MTT S5000

[baominghelly]

Superseded by #42, which has the identical tested Git tree in a single linear commit so it can satisfy this repository's rebase-only merge policy.