Install
CUDA & Compute
CUDA, creator workloads, AI frameworks, and encoder/decoder blocks.
- 17 Tracked terms
- Last 30 days Feed window
What this topic collects on
An article joins this feed when it matches these terms. Each one is also a search of its own.
Related topics
Latest in CUDA & Compute
The 552B DeepSeek V4.1-Flash Model Offers A Peak Output Of 494 Tokens/Second When Powered By An At-Home Rig Spanning 4x NVIDIA DGX Spark Units
10+ hour, 57+ min ago (241+ words) The AMD veteran and tech expert, Patrick Moorhead, has just disclosed his at-home AI setup, which consists of 4x NVIDIA DGX Spark units - offering a pooled unified memory of ~512 GB - stacked inside a compact "tec MOJO" rack with additional cooling fans…...
MGLRU-FG Delivering Up To 10~40% Higher Performance For Linux In Some Tests
1+ day, 12+ hour ago (206+ words) Linux's Multi-Gen LRU "MGLRU" has already achieved nice performance gains for Linux in recent years by optimizing page reclaims and improves performance under memory pressure. The new MGLRU-FG patches aim to take things further... - Categories Computers Display Drivers Graphics Cards…...
I Ran Gemma 4, Qwen3–8B, and GPT-OSS Locally. Here’s What I Found
1+ day, 9+ hour ago (160+ words) I used to treat local AI as a hobby for people with spare GPUs and too much free time. Then I got tired of sending every small prompt to a paid API and wondering where my data went. So I…...
Training a neural network on the GPU in pure PHP (and what kernel fusion did for it)
1+ day, 16+ hour ago (411+ words) PHP is not where anyone expects GPU computing to happen. I wanted to find out how far it can go without a Python sidecar, so I built php-gpu-tensors, a native PHP extension for NVIDIA GPUs. This post walks through the…...
Can You Run Gemma 4 12B or E4B Locally? RAM, VRAM and Mac Requirements
1+ day, 18+ hour ago (676+ words) For a first local text-chat setup, start with Gemma 4 E4B on a 16GB machine, or Gemma 4 12B on a 24GB–32GB machine. Those are conservative buying and setup recommendations based on the exact four-bit files below and explicit headroom allowances. They are not measured minimums....
DataFusion on the Apple Silicon GPU: One Optimizer Rule, the Same SQL, Sorts 6.9x to 28.8x Faster
3+ day, 5+ hour ago (32+ words) Run Apache DataFusion on the Apple Silicon GPU: ArrowMetal 0.4.0 adds one optimizer rule. Same SQL, DataFusion's answers, full sorts 6.9x to 28.8x fas. Tagged with rust, datafusion, gpu, apple....
gemma-4-31b-it-oQ8e-mtp on M4 Pro (16c) — oMLX Benchmark
4+ day, 23+ hour ago (19+ words) gemma-4-31b-it-oQ8e-mtp on M4 Pro (16c) oMLX Up to 16 results from this session are shown, including the current result....
gemma-4-26b-a4b-it-oQ8e-mtp on M4 Pro (16c) — oMLX Benchmark
4+ day, 23+ hour ago (19+ words) gemma-4-26b-a4b-it-oQ8e-mtp on M4 Pro (16c) oMLX Up to 16 results from this session are shown, including the current result....
Gemma 4 on Amazon SageMaker or a VM? The Same Model Server at 1.40x the Price
5+ day, 9+ hour ago (562+ words) This article serves the same Gemma 4 build from an Amazon SageMaker endpoint and from a plain EC2 instance with the same GPU, on an NVIDIA T4 and an NVIDIA L4, and prices both against Compute Engine and Cloud Run. A suite of Python…...
PQC-Bench Part 1 — measuring X25519MLKEM768 costs on a $0.04/hour Graviton3
1+ week, 7+ hour ago (552+ words) Complete setup — Terraform, Gatling simulation, driver script, raw Gatling result files — is in the repo. This can be reproduced in around ~50 minutes for under $0.30 (based on my experience, accounting for the occasional disruption during a test). The server serves an…...