François Chollet:AI 算力需求未触顶,当前是罕见的买入机会
Keras 作者 Chollet 发了条观点帖,用 HBM 售罄到 2029 年、训练算力 5 个月翻倍、推理 70-85% 利润率这些数据反驳泡沫论。正在纠结要不要囤算力、投算力的人可以看看他的逻辑。
François Chollet 在 X 上发文表态:要么认为 AI 建设浪潮已经结束、算力需求见顶,要么现在就是罕见的买入机会。他给出三组数据支撑:HBM 和 DRAM 产能到 2029 年已被售罄,芯片和内存供应预计紧张到 2030 年代;前沿模型训练算力约每 5 个月翻一倍,模型能力提升速度是推理模型出现前的两倍以上。
他还指出推理业务目前有 70-85% 的利润率,需求呈几何级增长,且每一代新模型都会解锁新的应用场景。GPU 现货价格持续上涨,算力仍难以获得。
Chollet 强调自己并非 AI 多头,只是基于供需现实作出判断。这一表态针对的是当前市场上"AI 基建泡沫论"的分歧,对关注算力投资和采购节奏的开发者与企业有直接参考意义。
Either you believe the AI buildout is over, that AI was a flash in the pan, and that the need for compute has already peaked... or this a rare buying opportunity.
Reminder: GPU spot prices keep rising and it's hard to secure capacity. HBM and DRAM production is effectively sold out through 2029, with chip and memory supply projected to stay tight until well into the 2030s.
AI research isn't slowing down: frontier training compute is doubling every ~5 months, and model capabilities are scaling over twice as fast as they did pre-reasoning. Inference is profitable with 70-85% margins, and demand is growing geometrically. Each new model generation unlocks new use cases.
I'm not even an AI booster, just being realistic here.
本条情报归属 AITOP「行业」核心栏目,重点聚焦北美前沿实验室与硅谷科技巨头。涉及核心实体与技术议题包括:François Chollet、GPU、HBM、算力,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明:本文原始线索采编自「François Chollet」(@fchollet),经 AITOP 多维评分管线结构化提炼。