用于AI的GPU的快速折旧是一个重大担忧，其寿命估计不到三年。这种短暂的运行寿命，加上这些组件的高资本支出和能源消耗，引发了对其利用率和价值的疑问，一旦当前的AI热潮消退。Core Weave和Microsoft等实体对GPU的所有权进行整合，在转移风险的同时，也集中了潜在的未来负债和环境问题。 AI
影响 GPU寿命短和高能耗引发了对未来数据中心废弃和环境影响的担忧。
排序理由 该集群讨论了AI中GPU使用的经济和环境影响，重点关注折旧和潜在的未来浪费，这属于对行业的评论。
在 Mastodon — fosstodon.org 阅读 →
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
@ carnage4life maybe, but GPU parts are high CAPEX items with MW consumption and no use outside machine learning and scientific workfloads. When the AI bubble collapses those parts still need to be utilised enough to cover their depreciation and at a spot price which also covers …
@ ChrisMayLA6 The lifespan of the GPUs used for # AI is less than three years. IMO, scaling and oligarchical behaviour mean that replacements will be consolidated, not distributed (likely where the poor live and the politicians give generous tax breaks to polluters) This means, b…