Types gt::bfloat16_t and gt::complex_bfloat16_t

[cmpfeil]

The types gt::bfloat16_t and gt::complex_bfloat16_t are provided as 1:1 copies of gt::float16_t and gt::complex_float16_t, respectively.

The 1:1 adaptions from [complex_]float16_t.h --> [complex_]bfloat16_t.h are:

• cuda_fp16.h --> cuda_bf16.h
• __half --> __nv_bfloat16
• CUDA_ARCH >= 530 --> CUDA_ARCH >= 800
• GTENSOR_ENABLE_FP16 --> GTENSOR_ENABLE_BF16

Similarly, test_[complex_]bfloat16_t.cxx is a copy of test_[complex]float16_t.cxx with any instance of float16_t replaced by bfloat16_t.

Tested on A100, built with modules gcc/11 and cuda/11.4 loaded as well as -DGTENSOR_DEVICE=cuda -DGTENSOR_ENABLE_BF16=ON -DCMAKE_CUDA_ARCHITECTURES=80 added to the cmake call.

Note: Obviously, it would be cleaner to have a class templated on storage_type and compute_type covering both, gt::float16_t and gt::bfloat16_t, at the same time (and possibly also a future gt::float8_t). However, as this is only a temporary solution until NVIDIA supports this within thrust, it might be good enough. Any opinions on this?

Any feedback is highly appreciated!

[bd4]

This looks fine to me. Some day we might have official clean support for all these types in CUDA/HIP/SYCL, but in the meantime just having something that works is very useful and I am not too concerned about making it pretty.

[germasch]

Other than the _ for private member variables, looks good to me. That is probably the case for float16_t as well (haven't checked), so could be updated in a subsequent PR.

The other comment I have is that it may be worth checking whether the duplication of the tests could be avoided (googletest does supported templated tests, though we're generally not taking advantage of that, but we probably should...)

[germasch]

We generally use an underscore suffix for private member variables, so x_ here would be appreciated, as it serves as a hint that something is a member variable when it's used within class member functions, and it arguably makes x(x) in the constructor a bit less confusing.