西风 发自 凹非寺 量子位 | 公众号 QbitAI 马斯克"巨硬计划"新消息，第三栋专属厂房已经买下来了，代号MACROHARDRR。 果然更硬核，老马透露，其将具备2GW供电规模。 按照规划，Colossus II第一阶段部署11万个英伟达GB200 GPU，最终目标是超过55万个GPU，峰值功率需求预计超过1.1GW。 马斯克巨硬（Macrohard）项目，也是在8月份公开的。 后续，马斯克承认Colossus II正是巨硬计划的一部分。 什么概念？ 1GW电力足以满足约75万户美国家庭的用电需求。 若参照此前曝光的（200MW支持11万台GB200）的功耗密度与架构效率推算，可支持约110万台英伟达GB200 NVL72 GPU。 到建成，xAI算力储备将再创全球之最。 算力巨兽持续扩容 马斯克口中第一厂，早已为人所熟知——Colossus I。 Colossus I验证了xAI大规模自建超级计算集群的可行性，从无到有建成所有配套设施仅用了122天，"从第一个机架落地到开始训练任务，只用了19天"。 至今，Colossus I仍是全球规模最大、运行状态最稳定的单一算力集群，配备约20万颗英伟达H100 ...

西风 发自 凹非寺 量子位 | 公众号 QbitAI 马斯克"巨硬计划"新消息，第三栋专属厂房已经买下来了，代号 MACROHARD RR 。 果然更硬核，老马透露，其将具备 2GW 供电规模。 什么概念？ 1GW电力足以满足约75万户美国家庭的用电需求。 按照规划，Colossus II第一阶段部署11万个英伟达GB200 GPU，最终目标是超过55万个GPU，峰值功率需求预计超过1.1GW。 马斯克巨硬（Macrohard）项目，也是在8月份公开的。 若参照此前曝光的（200MW支持11万台GB200）的功耗密度与架构效率推算， 可支持约110万台英伟达GB200 NVL72 GPU 。 到建成，xAI算力储备将再创全球之最。 算力巨兽持续扩容 马斯克口中第一厂，早已为人所熟知—— Colossus I 。 Colossus I验证了xAI大规模自建超级计算集群的可行性， 从无到有建成 所有配套设施仅用了122天 ，"从第一个机架落地到开始训练任务， 只用了19天"。 至今，Colossus I仍是全球规模最大、运行状态最稳定的单一算力集群，配备约20万颗英伟达H100/H200和约3万颗英伟达GB200 ...

Core Insights - Elon Musk's "Macrohard" initiative has rapidly established a computing cluster capable of supporting 110,000 NVIDIA GB200 GPUs within six months, achieving a power supply scale of 200MW, which is a record compared to similar projects by OpenAI and Oracle that took 15 months [1][2][4] Group 1: Project Overview - The "Macrohard" project, which started in 2021, aims to automate the entire software development lifecycle using AI agents, including coding, design, testing, and management [2][4] - The Colossus II project, initiated on March 7, 2025, plans to deploy over 550,000 GPUs, with a peak power demand expected to exceed 1.1GW, and a long-term goal of expanding to 1 million GPUs [4][5] Group 2: Infrastructure and Power Supply - To meet the substantial power requirements, xAI has acquired a former Duke Energy power plant in Mississippi, which has been temporarily approved to operate gas turbines for 12 months [4][5] - xAI has partnered with Solaris Energy Infrastructure to lease gas turbines, with 400MW currently allocated to the project, and has invested $112 million in capital expenditures for this partnership [5] Group 3: Strategic Importance - The Macrohard initiative is becoming a crucial part of Musk's business strategy, positioning Tesla as an "AI robotics company," with 80% of its future value tied to robotics [6] - The AI software developed through Macrohard will enhance Tesla's autonomous driving algorithms and factory automation, while Tesla's extensive real-world data will provide valuable training data for the Macrohard project [6]

Core Insights - Musk's "Macrohard" initiative aims to build a powerful computing cluster, achieving a 200MW power supply capable of supporting 110,000 NVIDIA GB200 GPUs NVL72 in just six months [1][12] - The project has outperformed collaborations between OpenAI and Oracle, completing in six months what took them 15 months [2] - The Colossus II computing cluster is designed to automate the entire software development lifecycle using AI agents, simulating a complete software development team [3][5] Group 1 - Colossus II project was initiated on March 7, 2025, with xAI acquiring a 1 million square foot warehouse and adjacent land totaling 100 acres in Memphis [10] - The first phase of Colossus II aims to deploy 110,000 NVIDIA GB200 GPUs, with a long-term goal of exceeding 550,000 GPUs and peak power demand expected to surpass 1.1 gigawatts [13][14] - To meet the substantial power requirements, xAI has adopted a cross-regional energy strategy, acquiring a former Duke Energy power plant in Mississippi to operate gas turbines [15] Group 2 - The project is currently in a critical phase, with Musk personally overseeing operations and maintaining a rigorous schedule to ensure progress [16] - Tesla's positioning as an "AI robotics company" indicates that 80% of its future value will derive from robotics, with Macrohard's AI software enhancing Tesla's autonomous driving algorithms and factory automation [17]