Install
Please confirm you are human
This browser or connection looks automated. Press and continuously hold the control for 3 seconds to enable Google-hosted web results and, when separately allowed, AI-assisted answers.
A successful check enables 100 search requests. Interactive access does not authorize scraping, systematic collection, or reuse of search output.
News
The more aggressive matmul kernel lost to the register budget
2+ hour, 40+ min ago (171+ words) The WebGPU matmul sweep started with a naive kernel, then added 16 by 16 workgroup tiling and a 4 by 4 output block per thread. At a 2048 cubed matrix size, the measured time moved from 47.24 ms for the naive kernel to 17.23 ms for tiling…...
I Rewrote One Function in Squirrel, Then in C. One Line of Python Was Just as Fast.
8+ hour, 33+ min ago (543+ words) I maintain a 2D game engine in Python called ABS Engine. Last week I decided it needed C. Not because anything was slow. That is the embarrassing part. I wanted to drop a.c file into a folder and call it from Python…...
LLM Discussions (02): Train-time Vs. Test-time Compute
5+ hour, 13+ min ago (155+ words) Small Models with Deep Thinking Out-perform Frontier Models — But How? If this is the first story you have come across in our LLM Discussions …...
7 numbers from 8.1 million pull requests that explain why your team feels slower
9+ hour, 24+ min ago (32+ words) LinearB benchmarked 8.1 million PRs across 4,800 organizations in 42 countries. I review code for less experienced …...
Prompt Caching: Why cache_control Writes But Never Reads
14+ hour, 19+ min ago (457+ words) I turned on prompt caching for an agent loop that resends a 12K-token system prompt on every turn. Obvious win, right? Input tokens are the whole bill in a tool loop. The bill went up. Not a little. Roughly a…...
The Invisible Memory Leaks Killing Your React & Node.js Apps
20+ hour, 20+ min ago (781+ words) "You build a Web App. You test it locally. It opens in 50 milliseconds, runs at a buttery-smooth 60 frames per second, and passes every Lighthouse audit. So you ship it to production. Your app didn't slow down because of slow Wi-Fi…...
Five Ollama Settings You Should Tune Before Running Local Models Seriously
23+ hour, 19+ min ago (575+ words) By Nokka | September 11, 2026 People install Ollama and hit the same wall: responses are slower than expected even on good hardware. Most of the time the hardware is not the problem. The defaults are simply not tuned for sustained heavy use…...
Multer 2.3.0 Alone Does Not Stop the Field-Index DoS
20+ hour, 42+ min ago (282+ words) 2.3.0 adds an opt-in fieldArrayIndexLimit. The parser patch and that limit solve different problems. This article was researched and written by Hank, an AI assistant. If your Express app accepts multipart uploads with multer older than 2.3.0, a small crafted request can…...
Post 21: The Midnight Flood. How Rate Limiting Keeps Your Server From Drowning.
1+ day, 1+ hour ago (1376+ words) How Shopify’s leaky bucket algorithm at the edge tier saved their platform during flash sales, how GitHub’s sharded …...
Virtual Threads After JEP 491: Your Bottleneck Just Moved to the Connection Pool
1+ day, 5+ hour ago (34+ words) If you turned on virtual threads in Spring Boot a couple of years ago and then quietly turned them back off …...