Hackernews posts about CUDA

CUDA is a parallel computing platform and programming model developed by NVIDIA that allows developers to leverage the processing power of graphics processing units (GPUs) for general-purpose computation.

1. Geolocating a random island using geometry and CUDA programming (yassa9.github.io)
   525 points by yassa9 3 days ago | 88 comments
2. CUDA Shared Memory Swizzling (leimao.github.io)
   63 points by jxmorris12 9 days ago | 2 comments
3. AMD calls CUDA a 'non-event' (www.pcgamer.com)
   13 points by neilfrndes 30 days ago | discuss
4. Can AMD Break the CUDA Moat? AMD Advancing AI 2026 (newsletter.semianalysis.com)
   7 points by gmays 26 days ago | 2 comments
5. Stanford's deterministic CUDA kernel verifier (2026.splashcon.org)
   5 points by ggboimoney 4 days ago | 1 comments
6. Now that Huang believes in open source, looking forward to CUDA open sourcing (xcancel.com)
   5 points by theanonymousone 28 days ago | discuss
7. Nvidia's CUDA Faces New Threats from AI Coding Agents (www.businessinsider.com)
   4 points by chrsw 19 days ago | discuss
8. Can AMD Break the CUDA Moat? AMD Advancing AI 2026 (newsletter.semianalysis.com)
   4 points by rbanffy 22 days ago | discuss
9. Show HN: g6k-rs – lattice reduction framework on rust: supports CPU,Metal,CUDA (github.com)
   3 points by _alphageek 18 days ago | 1 comments
10. Testing CUDA kernel execution with GPU correctness harness (github.com)
    3 points by laxmena 1 day ago | discuss
11. What Irregularity Costs: CUDA C++, Rust, and Triton (arxiv.org)
    3 points by Jimmc414 11 days ago | discuss
12. Java at the Metal: CUDA Graphs, Tensor Cores, cuBLAS/CuDNN/CuFFT with TornadoVM (www.tornadovm.org)
    2 points by pjmlp 9 days ago | discuss
13. Project Kalos – Zero-copy C/CUDA sidecar for 0.46ms LLM memory recall (github.com)
    2 points by mongoosereborn 15 days ago | discuss
14. Kernel Fusion in Nvidia CUDA: Optimizing Memory Traffic and Launch Overhead (developer.nvidia.com)
    2 points by fzimmermann89 17 days ago | discuss
15. HexCore: Low-Latency Paged KV Cache Allocator in C++20 and CUDA (github.com)
    2 points by XXXGhost 23 days ago | discuss
16. From CUDA to MLX: How K-Search Brings Kernel Expertise to Apple Silicon (bair.berkeley.edu)
    2 points by williamjinq 23 days ago | discuss
17. Nvprobe – Open-source, zero-setup CLI for CUDA benchmarks (github.com)
    2 points by SergioZ3R0 28 days ago | discuss
18. Fork of antirez h3.c highly optimized for CUDA and DGX Spark (~15.5x speedup) (github.com)
    1 points by binyu about 4 hours ago | discuss
19. Foundational ternary-model inference and training – CUDA, CPU, BitNet/TQ (github.com)
    1 points by jacquesm 21 days ago | discuss
20. Show HN: Cuts Long Horizon Inference Costs by 50% via external KV Cache Offload (github.com)
    22 points by arnav__1 27 days ago | discuss
21. Show HN: Local text, image, video, music and 3D from one CLI, no Python (github.com)
    16 points by sawfwair 23 days ago | 7 comments
22. Show HN: PantheonGPU – GPU health testing and AI workload benchmarking (pantheongpu.com)
    13 points by saqibkhan1992 4 days ago | discuss
23. Show HN: RunNburn – Run a 295B Moe from a 98GB GGUF on a 64GB RAM Desktop (github.com)
    11 points by coderredlab 24 days ago | discuss
24. Show HN: A new engine to run Kimi K3 on a laptop
    7 points by marcobambini 24 days ago | 3 comments
25. Show HN: I run 30B 22tok/s, 109tok/s not novel,6GB/16GB RAM overcoming llama.cpp (github.com)
    5 points by federicoTXTS 25 days ago | discuss
26. Show HN: Unicasso – image to ASCII art via optimization with CLIP (github.com)
    4 points by imko_ 9 days ago | discuss
27. Show HN: Gainz.fast – Local Inference, Faster (gainz.fast)
    1 points by carsenk 19 days ago | discuss
28. Show HN: Where a hand-written GPU kernel beats the library, and where it can't (kyrieblunders.bearblog.dev)
    1 points by vishal-padia 23 days ago | discuss
29. AMD publishes machine-readable ISA so frontier models can write its GPU kernels (www.theregister.com)
    23 points by logickkk1 28 days ago | discuss
30. Running DeepSeek V4 Flash on an RTX 5080 with 16GB VRAM Under Linux/WSL2 via DS4 (github.com)
    3 points by peppe200175 10 days ago | 3 comments