Comments (5)
That depends on your targeted toolchain. Maybe this is a good start:
https://forums.developer.nvidia.com/c/accelerated-computing/hpc-compilers/nvc-nvc-and-nvfortran/313
from libcudacxx.
@jrhemstad
https://forums.developer.nvidia.com/t/feature-request-fp32-multiplication-with-zero-optimization-specifier/245926/5?u=saitamatenthousand
from libcudacxx.
Hi @Saitama10000 this is not a library feature but a compiler feature.
As library, we have no say in how the compiler interprets our code. So please file this feature request with the compiler toolchains
from libcudacxx.
Hi @miscco I'm having a hard time finding where I should file this request. Can you help me out with a link?
Thank you,
Saitama10000
from libcudacxx.
@Saitama10000 what optimization are you hoping to achieve? If the coefficients are known at compile time and your function is inlined, it will exploit the fact that the values are known at compile time.
Example: https://godbolt.org/z/K6Khrb446
Even if the values aren't known at compile time, there is no difference in the type of FMA
that is generated.
from libcudacxx.
Related Issues (20)
- [QUESTION] Can we remove the `isnan` check in the `complex<T>` multiplication operator? HOT 6
- [Question] understanding documentation of cuda::atomic::atomic_thread_fence HOT 2
- CUDA Tuple Structured Binding Declaration Broken HOT 4
- libcu++ should document policy for mixing a custom version with the CUDA Toolkit HOT 2
- `error: "cuda::std::cuda" is ambiguous` when including <cuda/std> HOT 6
- Issue with cuda::atomic_ref::notify_one and wait HOT 1
- Copyright in `nv` files HOT 2
- A device variable cannot be marked constexpr (cuda::std::swap / CTK 11.2) HOT 2
- `proclaim_return_type` and `__invoke` should be decorated with `nv_exec_check_disable` HOT 2
- Add `cuda::std::expected`
- Placement new creates performance issues
- libcu++ should not define __host__/__device__/__forceinline__ (or any other runtime macros) HOT 2
- atomic fence questions HOT 2
- Hide if target bits from NVCC/NVRTC
- Function objects in `<functional>` need to be annotated with `nv_exec_check_disable` HOT 1
- Does `cuda::std::complex` have opt-in for simplified complex math HOT 3
- libcu++ should build with gcc-6
- Passing `cuda::proclaim_return_type` to `thrust::make_transform_iterator` gives compilation error HOT 1
- Support fixed-width floating point types
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from libcudacxx.