金粉世家
Chinese firm delivers self‑developed core casting for world's most powerful onshore wind turbine_我的网站
一 |
The core castings for the world's most powerful onshore wind turbine in terms of single‑unit capacity have recently been delivered by a Chinese company, marking a technological breakthrough in the large‑megawatt onshore wind power castings, domestic news portal stdaily.com affiliated with Science and Technology Daily reported on Wednesday.
Independently developed by Dalian Huarui Heavy Industry Group Co., Ltd. (DHHI), the delivered components include the bearing housing and base castings for 15 MW doubly fed onshore wind turbines. The components need to bear the huge gravity, wind loads and alternating loads transmitted from the blades and hub, and transfer the loads to the generator, ensuring steady rotation of the main shaft.
The products passed all customer inspections and assessments in one go and were officially delivery, stdaily.com reported.
Adopting an advanced modular design, the turbine components delivered can support flexible power adjustment ranging from 13 MW to 16 MW, enabling adaptive operation to varying and complex wind conditions across diverse regions, the report said.
The turbine's rotor diameter ranges from 240 meters to 260 meters, and are compatible with tower heights of 135 meters to 190 meters, delivering a maximum swept area of 53,000 square meters, roughly 8 standard football pitchs (105m×68m). This places the unit's overall performance among the world's top-tier.
Such super-large castings place extremely high demand on material properties, casting processes, dimensional accuracy and internal quality. The successful validation and delivery of the products therefore demonstrate the research and development, as well as manufacturing capabilities in super‑large wind power castings, stdaily.com reported.
Global Times
。
在AI大模型训练与推理需求爆发的推动下，AI芯片行业正经历深刻重构：竞争焦点从单一制程性能转向“专用架构+软件生态+能效效率”的系统级较量。而OpenAI自研AI芯片的最新消息，无疑泛起了新的涟漪。 据财联社报道，OpenAI自研的推理芯片Jalapeno迎来了新的突破，在测试中的表现优于英伟达当前的产品阵容。 OpenAI芯片业务负责人Richard Ho在接受采访时表示，Jalapeno芯片与英伟达GB300进行的比较测试中，Jalapeno在两个指标上占据优势：单位功耗所能处理的AI工作量，以及响应的速度。 需要指出的是，Jalapeno并未与英伟达最新一代的Vera Rubin芯片进行测试，后者刚刚开始出货。
二 | 此外，Jalapeno也并非为AI模型训练而设计，其主要面向AI推理阶段，即模型完成训练后，根据用户提示生成响应并执行任务的阶段。而训练则是英伟达技术最擅长的领域之一。 公开资料显示，除了英伟达这一行业领跑者之外，OpenAI等全球科技巨头均在积极自研AI芯片，比如亚马逊自研的Trainium系列、Inferentia系列芯片，微软自研的Maia系列芯片，Meta自研的MTIA系列芯片，特斯拉自研的Dojo D1、AI5等芯片，谷歌自研的TPU系列芯片。 国内龙头同样在积极自研AI芯片，比如阿里巴巴的含光、真武系列，百度的昆仑芯，摩尔线程的长江、华山等，快手的SL200，以及字节跳动的SeedChip等等。
三 | 头豹研究院表示，巨头们积极自研AI芯片，背后的深层次原因在于美国持续收紧先进芯片及制造设备出口管制，叠加全球AI资本开支向头部云厂商高度集中，加速了算力供应链的地缘割裂，各国纷纷押注自主生态。 亿欧智库表示，当前国产替代已告别零散的技术突破，进入全产业链协同攻坚的新阶段，AI芯片已完成“设计制造-设备-材料”全链条布局。根据三方机构预测，到2030年，中国AI芯片自给率将升至86%，市场规模突破670亿美元。 从产业链来看，亿欧智库表示，可关注EDA/IP（华大九天、概伦电子、广立微等）、设备（盛美上海、中微公司、华海清科等）、芯片设计（寒武纪、海光信息、云天励飞等）、封装测试（通富微电、长电科技、华天科技等）等核心环节。
四 | 
（文章来源：东方财富研究中心）。
五 |
Current article:http://8ho2z4.ruwentaqintiaozhupapaigai.shop/news/20260826_604373.html
Published on:22:08:35