#GPUcomputing
A primer on GPUs and kernel optimization
Discussion | hackernews | Author: nomi_uw

#GPUComputing
A primer on GPUs and kernel optimization
www.nabo.dev
November 9, 2026 at 7:08 PM
NVIDIA released CUDA 13.5 with general availability for Vera Rubin. The toolkit also includes CCCL 3.5. Existing application speed still needs testing.

news.s3dat.com/stories/nvid...

#NVIDIA #GPUComputing

AI-generated video.
November 5, 2026 at 2:23 AM
Six GPUs on One X16 Slot: R9700s and RTX 3090s Behind a PCIe Switch
Discussion | hackernews | Author: dougcalobrisi

#GPUComputing
Six GPUs on One X16 Slot: R9700s and RTX 3090s Behind a PCIe Switch
Four Radeon AI PRO R9700s and two RTX 3090s behind a PEX88096 switch on one X570 slot: BIOS, kernel flags, ROCm and CUDA, and Qwen3.8-27B at 601 tok/s.
doug.sh
November 9, 2026 at 9:08 AM
🚀 HPC Explained #7: What is GPU computing?
GPUs can perform many similar calculations simultaneously, making them highly useful for HPC.
They accelerate workloads such as scientific simulations, AI, and data analysis through massive parallelism. 💡
#HPCExplained #HPC #GPUComputing #Supercomputing
November 3, 2026 at 9:45 AM
Speeding up GPU copies by permuting iteration spaces, or, optimising the identity function
Discussion | lobsters | Author: fanf

#GPUComputing
Speeding up GPU copies by permuting iteration spaces, or, optimising the identity function
A high-performance and high-level purely functional data-parallel array programming language that can execute on the GPU and CPU.
futhark-lang.org
October 31, 2026 at 5:08 PM
Meta Andromeda: Powering the Next Generation of Ads Retrieval Through Model-System Co-Design
tech_blogs_facebook

#GPUComputing
Meta Andromeda: Powering the Next Generation of Ads Retrieval Through Model-System Co-Design
This post describes the next generation of Meta Andromeda, Meta’s personalized ads retrieval engine, and how co-designing the model, system, and GPU hardware unlocks a new efficiency frontier…
engineering.fb.com
October 28, 2026 at 4:30 PM
GPU-unlock: FP/BF16 instruction issue rate on Nvidia 3000's cards
Discussion | hackernews | Author: fragebogen

#GPUComputing
GPU-unlock: FP/BF16 instruction issue rate on Nvidia 3000's cards
Unlock FP/BF16 instruction issue rate on Nvidia 3000's cards to their professional equivalents - Zagitta/gpu-unlock
github.com
October 24, 2026 at 2:07 PM