Trending storyDeveloping
PROMETHEVS runs DeepSeek locally on a 12 GB GPU
1 report1 sourceupdated 2 hours ago
What happened
AI digest
A Hacker News front-page post on November 7 describes PROMETHEVS 0.2, a local inference engine that runs DeepSeek-V4.1-Flash-1017 on a machine with 64 GB of RAM and an RTX 3060 with 12 GB of VRAM. The engine runs the model with compression followed by error correction, and builds its store from all 510 GB of the original weights. The 12 GB is graphics memory; the machine also has 64 GB of system RAM.
Written by AI from the coverage · updated 53 minutes ago
Coverage
Follow the reports to see the story from different sides.
Nov 7
- Hacker News front pagePromethevs: Full DeepSeek-v4.1-Flash-1017, locally, on 12GB of VRAM
PROMETHEVS 0.2 提供本地推理引擎，在 64 GB RAM、RTX 3060 12 GB 的电脑上以压缩后纠错方式运行 DeepSeek-V4.1-Flash-1017，使用全部 510 GB 原始权重构建存储。
Heat over time
Not enough continuous observations to draw a trend yet.