炒股就看金麒麟分析师研报，权威，专业，及时，全面，助您挖掘潜力主题机会！
来源：清爽AI报
英特尔CEO陈立武警示，AI发展的关键瓶颈已从算力转向内存。大模型训练与推理过程中，GPU与内存之间每秒需交换数十TB级数据，如同顶级跑车受困于纤细油管。全球能大规模量产高端AI所需HBM内存的仅有三家企业，且产能已被预订至2028年，导致算力与内存扩产严重失衡。此矛盾在三个层面引发系统性卡点：短期导致算力利用率低下，推高模型训练与推理成本；中期将考验芯片制造产能与半导体工业体系承载极限；长期则挑战电力、散热、材料、互联等基础设施支撑能力。AI竞争已超越算法层面，演变为对国家级完整产业体系——涵盖算力、内存制造、能源与基础设施——的长期耐力比拼。未来决定AI上限的，并非偶发性的模型突破，而是支撑复杂系统持续稳定运转的深厚工业底蕴。模型与框架会迭代，但芯片、内存、材料等底层能力一旦落后，则需漫长周期追赶。AI的终极形态是稳定运转的智能工业机器，而非单一的聪明模型。
责任编辑：杨赐
Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.