TAN(3clc)

πŸ“¦
version 1.0~svn33624 Β· 3 versions
, in ubuntu jammy, ubuntu noble

✨ NAME ✨

tan, _tanh, _tanpi, _half_tan, _native_tan β€” the tangent functions, bestie πŸ’– (yes the whole squad)

gentype tan(gentype );

gentype tanh(gentype );

gentype tanpi(gentype x);

gentype half_tan(gentype x);

gentype native_tan(gentype x);

πŸŽ€ DESCRIPTION πŸŽ€

  • tan β€” computes the tangent. that's it, that's the function πŸ’…
  • tanh β€” computes the hyperbolic tangent. different vibe, still iconic 🌊
  • tanpi β€” computes tan(pi * x). no cap, that's literally the whole thing.
  • half_tan β€” computes tangent, but plot twist: x must be in the range -216... +216. stay in your lane, x 😭
  • native_tan β€” computes tangent over an implementation-defined range. the maximum error is also implementation-defined. it's giving mystery, it's giving "ask your GPU vendor" πŸ¦‹

πŸ’… NOTES πŸ’…

okay so here's the lore you actually need πŸ‘‘

The vector versions of the math functions operate component-wise. The description is per-component. so like, each lane does its own lil thing, independently, slay.

The built-in math functions are not affected by the prevailing rounding mode in the calling environment, and they always return the same value as they would if called with the round to nearest even rounding mode. so don't come for them with your rounding-mode drama, they're unbothered ✨

The built-in math functions take scalar or vector arguments. For any specific use of these functions, the actual type has to be the same for all arguments and the return type β€” unless otherwise specified. consistency is the moment fr fr 🫢

The generic type name gentype is used to indicate that the function can take float, float2, float3, float4, float8, float16, double, double2, double3, double4, double8, or double16 as the type for the arguments. a whole buffet 🌈

If extended with cl_khr_fp16(3clc), the generic type name gentype may indicate half and half{2|3|4|8|16} as arguments and return values. half precision girlies rise πŸ‘‘

The generic type name gentypef is used to indicate that the function can take float, float2, float3, float4, float8, or float16 as the type for the arguments. floats only, no doubles allowed at this party πŸ’ƒ

The generic type name gentyped is used to indicate that the function can take double, double2, double3, double4, double8, or double16 as the type for the arguments. doubles only, the fancy side of the family πŸŽ€

Functions with the half_ prefix are implemented with a minimum of 10-bits of accuracy, i.e. an ULP value less than or equal to 8192 ulp. yes bestie, 8192. we love an honest queen πŸ’€

Functions with the native__ prefix may map to one or more native device instructions and will typically have better performance compared to the corresponding functions (without the native__ prefix). The accuracy (and in some cases the input range(s)) of these functions is implementation-defined. fast and loose and free, but read the fine print πŸ¦‹


πŸ“œ SPECIFICATION πŸ“œ

OpenCL Specification[1]


πŸ‘€ SEE ALSO πŸ‘€

mathFunctions(3clc) β€” go read the whole family, iykyk 🌸


🫢 AUTHORS 🫢

The Khronos Group. kings, honestly. πŸ‘‘


Copyright Β© 2007-2011 The Khronos Group Inc.

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and/or associated documentation files (the "Materials"), to deal in the Materials without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Materials, and to permit persons to whom the Materials are furnished to do so, subject to the condition that this copyright notice and permission notice shall be included in all copies or substantial portions of the Materials.

🚨 yes we have to include the whole legal paragraph, it's literally a condition of the permission. don't skip it, bestie, it's in the contract πŸ’€

πŸ“ NOTES (the footnote one, not the other one) πŸ“

  1. OpenCL Specification β€” page 244, section 6.12.2 - Math Functions. that's where the receipts live 🧾

and that's the whole page!! brought to you by The Khronos Group, dated 01/14/2021, in the OpenCL Manual, TAN, TANH, TANPI, HA(3clc). periodt. πŸ’–βœ¨πŸ¦‹

πŸ“¦ where this page comes from