本地运行 LLM 的开发者终于有了速度新选择——DiffusionGemma 的 4 倍加速意味着更流畅的交互体验，用 atomic[.]chat 的团队可以直接在单卡上体验，值得一试。
atomic[.]chat 在单块 H100（FP8）上对比了 DiffusionGemma（扩散文本模型）与 Gemma4 26B A4B（自回归模型）。DiffusionGemma 速度是 Gemma4 的 4 倍，改变了错误分布模式。自回归模型逐 token 生成，速度慢但精度高；扩散模型批量生成 token 再迭代修正，因此响应更快。这一对比展示了扩散模型在本地推理中的速度优势，可能改变未来本地 LLM 的部署选择。
atomic[.]chat shared a revealing comparison of loc…
atomic[.]chat shared a revealing comparison of local open-weight LLMs running on their own hardware.
They benchmarked the new DiffusionGemma (diffusion text model) vs. Gemma4 26B A4B (autoregressive model) on a single H100 (FP8).
The 4X speed of DiffusionGemma changes the shape of error.
- Autoregressive models move left to right, one token at a time, which is slower, but each new word is conditioned on the exact text already written.
- Diffusion models write many tokens at once, then revise the block over several passes, so they can feel fast because the model is not waiting to finish token 1 before starting token 2.
atomic[.]chat, a desktop app for running LLMs locally