The home for great developer writing.

We surface the best developer writing from thousands of independent blogs, updated daily.

Streamlining CUB with a Single-Call API

· NVIDIA Corporation · Jan. 21, 2026, 9:43 p.m.

C++ CUDA pytorch Simulation / Modeling / Design C++ api-design GPU programming software development

Summary

This blog discusses a new simplified approach to the CUB C++ template library, introducing a single-call API to enhance usability and streamline GPU primitive algorithm implementation, which could significantly benefit developers seeking efficiency.

Read full post on developer.nvidia.com →

MORE POSTS LIKE THIS

PRIK — Bring native Fortran and C code to Python

Fortran Lang Discourse · Aug 27, 2026

Python Fortran

There's always a niche: rawzip and the cost of ergonomics

Nickb · Jul 18, 2026

Rust zip library

Make Long-Running NVIDIA TensorRT Engine Builds Observable and Cancelable in Python or C++

NVIDIA Corporation · Jul 22, 2026

Agentic AI / Generative AI Build AI Agents

Build a Python library with a C++ extension

Python Developer Tooling Handbook – pydevtools.com · Jul 7, 2026

Python C++

Pystd standard library, similar-ish functionality with a fraction of the compile time

Martin Cracauer · Jun 26, 2026

C++ software development

CLRHack: Meta-object Protocol

Funcall Blogspot · May 31, 2026

dotnet ai-coding

Discover more posts →

AUTHOR

Sponsored

Zulip

Organized team chat for people who take work seriously. Topic-based threading keeps conversations focused.

Become a sponsor →