PulseAugur
中
实时 05:29:29
English(EN) @ bhg Thanks for sharing your thoughts. Here are the actual numbers Train Gemini Ultra ~150 GWh Distill Nano from Ultra ~1–5 GWh / ~1–3% One Gemini cloud query

Gemini Ultra 训练消耗 150 GWh；Nano 版本能耗可忽略不计

作者 PulseAugur 编辑部 · [1 个来源] ·

一位 Mastodon 用户分享了 Google Gemini 模型能耗数据，估计训练 Gemini Ultra 需要约 150 GWh。从 Ultra 蒸馏出更小的 'Nano' 版本估计消耗 1 至 5 GWh。单个 Gemini 云查询约消耗 0.24 Wh，而一年服务 Gemini（约每天 10 亿次查询）约消耗 90 GWh。 AI

影响 提供了对训练和运行大型 AI 模型相关重大能源成本的见解。

排序理由 该集群包含用户生成的 AI 模型能耗估算，而非官方发布或基准测试。[lever_c_demoted from research: ic=1 ai=0.7]

在 Mastodon — sigmoid.social 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

Gemini Ultra 训练消耗 150 GWh；Nano 版本能耗可忽略不计

本文如何被排名

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
该集群包含用户生成的 AI 模型能耗估算，而非官方发布或基准测试。[lever_c_demoted from research: ic=1 ai=0.7]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
infra, model release
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
112 days old
Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

  1. Mastodon — sigmoid.social TIER_1 English(EN) · [email protected] ·

    bhg 感谢分享您的想法。以下是实际数据：训练 Gemini Ultra 约 150 GWh，从 Ultra 蒸馏 Nano 约 1–5 GWh / 约 1–3%，一次 Gemini 云查询

    @ bhg Thanks for sharing your thoughts. Here are the actual numbers Train Gemini Ultra ~150 GWh Distill Nano from Ultra ~1–5 GWh / ~1–3% One Gemini cloud query 0.24 Wh / 0.00000016% 1 year of Gemini cloud serving (~1B queries/day) ~90 GWh / ~60% 👉One Nano on-device query ~0 datac…

    链接 troet.cafe/@bhg

相关实体

相关话题