Trending storyWatching
Run a 700B GLM on a Laptop: No GPU Needed, SSD as VRAM
1 report1 sourceupdated 3 days ago
What happened
Summary
一个GitHub项目火了：在没GPU的笔记本上跑700亿参数的GLM模型。办法是把SSD当显存用，用存储换速度。正文没披露具体延迟和精度损失，但思路很直接——内存不够硬盘凑。对没显卡的开发者来说，是个低成本试大模型的办法。
Coverage
Follow the reports to see the story from different sides.
Sep 26
- QbitAI · WeChatRun a 700B GLM on a Laptop: No GPU Needed, SSD as VRAM
A GitHub project goes viral: run a 700B-parameter GLM on a laptop without a GPU. The trick is using SSD as VRAM, trading storage for speed. The post doesn't disclose exact latency or precision loss, but the idea is straightforward: swap memory for disk. For developers without a GPU, this is a low-cost way to test large models.
Heat over time
Not enough continuous observations to draw a trend yet.