1256K context with 72MiB of KV cache on the GPU (opens in new tab)(github.com)GitHub3GreenGames9d ago0Save
2A 35B MoE on a 16 GB GPU, without the offload tax (opens in new tab)(lucebox.com)18GreenGames17d ago0Save
3PFlash: 10x prefill speedup over llama.cpp at 128K on a RTX 3090 (opens in new tab)(github.com)GitHub3GreenGames1mo ago1Save
4We got 207 tok/s with Qwen3.5-27B on an RTX 3090 (opens in new tab)(github.com)GitHub165GreenGames2mo ago52Save
5Show HN: OS Megakernel that match M5 Max Tok/w at 2x the Throughput on RTX 3090 (opens in new tab)(github.com)GitHub6GreenGames2mo ago1Save
6App-Use, Control Individual Applications with CUA Agents (opens in new tab)(trycua.com)2GreenGames1y ago1Save
7Show HN: Lumier – Run macOS VMs in a Docker (opens in new tab)(github.com)GitHub159GreenGames1y ago52Save
8Microsoft is reportedly about to lay off 3% of its workforce (opens in new tab)(techcrunch.com)13GreenGames1y ago2Save
9Polaris is giving free GPUs/CPUs for everyone (opens in new tab)(polariscloud.ai)3GreenGames1y ago0Save
10Improvements in reasoning AI models may slow down soon, analysis finds (opens in new tab)(techcrunch.com)3GreenGames1y ago0Save
11Microsoft and OpenAI are renegotiating their partnership (opens in new tab)(techcrunch.com)3GreenGames1y ago0Save
12Apple developing new chips for smart glasses, Macs, and more (opens in new tab)(techcrunch.com)6GreenGames1y ago0Save
13Show HN: Tlume – a CLI tool that converts Tart VM images for Lume (opens in new tab)(github.com)GitHub5GreenGames1y ago2Save
14Show HN: Lume – OS lightweight CLI for MacOS and Linux VMs on Apple Silicon (opens in new tab)(github.com)GitHub309GreenGames1y ago75Save
