Skip to content

Specialize numeric_limits for CUDA 12.8 FP types #3832

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
miscco merged 6 commits into from
Feb 17, 2025
Merged

Specialize numeric_limits for CUDA 12.8 FP types #3832

Show file tree
Hide file tree
Changes from 3 commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
64758be
implement limits for new fp types
davebayer Feb 17, 2025
4860b84
Merge branch 'main' into new_fp_limits
davebayer Feb 17, 2025
d8afc2d
modularize `numeric_limits`
davebayer Feb 17, 2025
89fc6d4
suppress unused parameter warnings
davebayer Feb 17, 2025
034489e
oops, suppress unused parameter warning everywhere
davebayer Feb 17, 2025
15d2dc7
its C4100, not C2200
davebayer Feb 17, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 12 additions & 0 deletions libcudacxx/include/cuda/std/__cccl/extended_data_types.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,6 +27,8 @@
#include <cuda/std/__cccl/preprocessor.h>

#define _CCCL_HAS_INT128() 0
#define _CCCL_HAS_NVFP4() 0
#define _CCCL_HAS_NVFP6() 0
#define _CCCL_HAS_NVFP8() 0
#define _CCCL_HAS_NVFP16() 0
#define _CCCL_HAS_NVBF16() 0
Expand Down Expand Up @@ -54,6 +56,16 @@
# define _CCCL_HAS_NVFP8() 1
#endif

#if _CCCL_HAS_INCLUDE(<cuda_fp6.h>) && _CCCL_HAS_NVFP8() && !defined(CCCL_DISABLE_NVFP6_SUPPORT)
# undef _CCCL_HAS_NVFP6
# define _CCCL_HAS_NVFP6() 1
#endif

#if _CCCL_HAS_INCLUDE(<cuda_fp4.h>) && _CCCL_HAS_NVFP6() && !defined(CCCL_DISABLE_NVFP4_SUPPORT)
# undef _CCCL_HAS_NVFP4
# define _CCCL_HAS_NVFP4() 1
#endif

// NVC++ supports float128 only in host code
#if !defined(CCCL_DISABLE_FLOAT128_SUPPORT) && _CCCL_OS(LINUX) \
&& ((_CCCL_COMPILER(NVRTC) && defined(__CUDACC_RTC_FLOAT128__)) /*NVRTC*/ \
Expand Down
Loading
Loading