DIFF.BLOG
New Following Discover Jobs
More
Top Writers Suggest a blog Upvotes plugin
Report bug Contact About Privacy
+
Create new post Create new blog
Add your blog
Sign up
Topics
Follow your own topics →
Menu
New Following Discover Jobs Top Writers
More
Suggest a blog Upvotes plugin Report bug Contact About
Sign up
Privacy Policy
The home for great developer writing.
We surface the best developer writing from thousands of independent blogs, updated daily.
Join Diff.blog
TOPICS

Advanced NVIDIA CUDA Kernel Optimization Techniques: Handwritten PTX

· NVIDIA Corporation · July 2, 2025, 9:07 p.m.
Data Center / Cloud Development & Optimization Simulation / Modeling / Design CUDA-X CUDA scientific-computing AI performance PTX
Summary
This blog post discusses advanced techniques for optimizing CUDA kernels using handwritten PTX, highlighting their importance in accelerating computing performance in AI and scientific applications. It aims to enhance understanding of GPU optimization among developers and engineers.
Read full post on developer.nvidia.com →
MORE POSTS LIKE THIS
cudamemcpy-tuner: cudaMemcpying up to 5x faster on a B300
Henry Zhu · Nov 15, 2026
CUDA benchmarking
NVIDIA Vera Rubin NVL72 Achieves up to 2x Blackwell Ultra Performance in MLPerf Training v6.1
NVIDIA Corporation · Nov 16, 2026
Agentic AI / Generative AI Data Center / Cloud
How I massively improved my AI inference performance without buying new hardware
Red Hat · Sep 23, 2026
Machine Learning Kubernetes
How to Create a GPU-Optimized Machine Image with HashiCorp Packer on GCP
freeCodeCamp.org · Apr 22, 2026
GPU VM Image
Claude is suddenly stupid, there is nothing you can do
Blog Kronis · Nov 14, 2026
user-experience software development
Coherent and exhaustive modern Fortran mathematical/scientific library
Fortran Lang Discourse · Nov 16, 2026
mathematics Fortran
Discover more posts →
AUTHOR
RECENT POSTS FROM THE AUTHOR
Choose how you want to continue.
Continue with GitHub Continue with Google
By continuing, you agree to our Privacy Policy.