2.1 Gigawatt Computing Power Aircraft Carrier Sets Sail! Microsoft Takes Over Texas AI Factory Project, Moves into Same Campus as OpenAI

marsbit发布于2026-03-30更新于2026-03-30

文章摘要

In the global AI computing race, Abilene, Texas, is emerging as a key strategic hub. Microsoft has officially taken over a large-scale data center construction project originally initiated by OpenAI. The collaboration, led by data center developer Crusoe, will see Microsoft and OpenAI as neighbors in the same campus. The project will feature two new state-of-the-art "AI factory" data centers, supported by a dedicated 900-megawatt power plant. With Microsoft’s involvement, the campus will expand to a total of 10 data center buildings, reaching a total computing capacity of 2.1 gigawatts. Notably, OpenAI has decided to halt further expansion at the Abilene site and will instead distribute new computing resources across other U.S. locations to optimize grid load and resource allocation. The campus already hosts major players like Microsoft, OpenAI, and Oracle. The 2.1-gigawatt scale signifies a shift from gigawatt-level to larger computing clusters, enabling improved energy sharing and reduced latency for AI model training. This transformation positions Abilene as a critical hub in global AI infrastructure, with physical proximity between Microsoft and OpenAI likely accelerating the development of next-generation large-scale AI models.

On the global map of AI computing power competition, Abilene, Texas is becoming a new high ground heavily invested in by tech giants.

On March 30, it was reported that Microsoft has officially taken over a large data center construction project in Abilene, Texas. The project was originally planned to be expanded by OpenAI, but now with Microsoft's involvement, these two deep allies in the AI field will become "neighbors" in the same computing power campus.

Powerful Alliance: Building an AI Factory with 900 Megawatt Support

This collaboration was initiated by the renowned data center developer Crusoe. Microsoft's entry will inject stronger momentum into the campus:

Core Infrastructure: The project will construct two state-of-the-art "AI factory" data center buildings, equipped with a dedicated power plant with a generation capacity of 900 megawatts.
Scale Leap: With the integration of Microsoft's project, the total number of AI data center buildings in the campus will increase to 10, and the overall computing power is expected to reach an astonishing 2.1 gigawatts.

Strategic Shift: OpenAI Turns to Nationwide Deployment

It is worth noting that behind this takeover is a subtle adjustment in OpenAI's infrastructure strategy:

Halting Expansion: OpenAI had previously decided not to further expand the existing scale of the Abilene campus.
Distributed Deployment: To optimize resource allocation and grid load, OpenAI plans to deploy its additional computing power resources to other regions in the United States.

Industry Significance: Computing Power Campuses Enter the "Gigawatt Era"

The campus has now gathered top players such as Microsoft, OpenAI, and Oracle. The 2.1-gigawatt computing power scale marks the evolution of hyperscale data centers from traditional "gigawatt" levels to even larger computing clusters.

Cluster Effect: Giants gathering in the same campus can not only share large-scale energy support but also physically reduce data exchange latency, improving the efficiency of large model training.

Conclusion: The "Intelligent Engine" on the Texas Plains

From a single expansion project to a gigawatt-level computing power aircraft carrier after Microsoft's takeover, Abilene is transforming from a Texas plain into one of the "hearts" of the global AI industry. When the parent company of Microsoft (China) Co., Ltd. and OpenAI become neighbors here, this physical proximity may预示 the faster birth of the next generation of hyperscale AI models.

你可能也喜欢

Block因强劲季度业绩上调2026年展望，称AI已触及几乎所有代码

金融科技公司Block公布强劲第二季度业绩后，上调了全年展望。季度毛利润同比增长25%至31.7亿美元，调整后每股收益1.02美元，均超预期。公司因此将全年毛利润指引上调至125.1亿美元，调整后营业利润指引上调至34.7亿美元。首席财务官表示，上调指引源于上半年的强劲执行力和下半年的发展势头。公司在股东信中披露，其智能体人工智能在6月已协助编写和审查了近全部生产代码变更。业务负责人称，相比年初，每位工程师的代码变更量提升了150%。此前，公司于2月进行了以AI为导向的重组并裁员4000人。

cointelegraph21分钟前

cointelegraph21分钟前

Mysten Labs 技术负责人加入 Anthropic，专注 AI 安全

Mysten Labs联合创始人兼首席技术官Sam Blackshear宣布将离职，加入人工智能公司Anthropic，专注于防御性安全研究。Blackshear是Sui区块链创建者Mysten Labs的创始团队成员之一。他表示，自己职业生涯中最快乐的是动手进行技术工作并探索新的问题领域，此次加入Anthropic正是这样一个机会。他选择在当前人工智能被日益用于识别软件漏洞、自动化钓鱼攻击和加速对区块链基础设施攻击的背景下投身安全领域。离职后，Blackshear仍将作为密切顾问参与Mysten及Sui生态，并继续参与Move基金会相关工作。Mysten Labs的联合创始人兼首席执行官Evan Cheng将接替他制定公司未来的技术愿景。

cointelegraph21分钟前

cointelegraph21分钟前

Michael Saylor：比特币跌一半，我的数字信贷在赚钱

在2026年6月12日的Bitcoin for Corporations圆桌讨论中，Michael Saylor阐述了在比特币熊市下的投资策略。他指出，尽管比特币价格下跌了约50%，但其公司推出的数字信贷产品（如STRC、SATA）仍能保持正回报，成功剥离了比特币约90%的波动率。Saylor强调，比特币作为“数字资本”年化波动率高达40%，难以被主流资金接受，因此需要创建由比特币支持、与法币挂钩且能产生收益的“数字信贷”和“数字货币”产品。这些产品并非取代比特币，而是与传统货币市场基金、稳定币等竞争，旨在将无法承受高波动的海量传统资本引入比特币生态。 Saylor详细解释了“数字货币”需具备的特性：与法币（如美元）挂钩、保持零波动、并能产生高于通胀的收益，这样才能被全球绝大多数资本接受。他澄清，数字信贷产品吸引的是原本配置于传统低收益资产（如货币市场基金）或寻求加密收益（如替代Tether）的资金，而非比特币持有者的资金，这实质上是扩大了比特币网络的资金基础。此外，Saylor回应了关于其公司（MicroStrategy）融资策略的争议。他辩称，公司通过发行股票（尤其是溢价发行）换取现金或比特币，只要交易价格高于每股净资产价值，就是增厚而非摊薄股东价值。他区分了债务与优先股等混合资本工具的不同，后者因赋予发行方更多选择权且永不到期，风险更低。Saylor坦言，评估比特币财库公司的价值（如mNAV指标）非常复杂，投资者必须基于对比特币未来价格和波动率的明确假设，建立完整模型，不能依赖单一指标。他认为这些新兴商业模式仍在快速演变中。

marsbit54分钟前

marsbit54分钟前

弄潮先溺，“加密 x AI”先驱 ai16z 的落幕

《弄潮先溺，“加密 x AI”先驱 ai16z 的落幕》 2024年底，ai16z 作为加密 AI 代理（Agent）概念的先驱项目，通过一个带有模因色彩的众筹（募资约7.5万美元）迅速崛起，三个月内市值飙升至26亿美元，并带动了整个AI Agent赛道估值接近百亿美元。当时，主流AI市场对“AI代理”的理解尚在概念阶段，加密市场已为其提前定价。然而，早期就有质疑指出，其核心的AI投资代理功能实为人工操作，但在市场狂热中并未引起重视。 2026年，真正的AI代理（如Claude、CodeX等）已广泛落地，实现了当初加密AI叙事中描绘的许多功能。但兑现这些价值的，是Anthropic、OpenAI等传统科技公司，其通过API调用和订阅服务等模式成功商业化。而绑定代币的加密AI项目，在代币价格崩盘后往往陷入开发者流失、社区维权、资金枯竭的困境。 ai16z项目创始人Shaw Walters近期宣布项目彻底关停，基金会解散且无任何回购。他表达了对加密社区的极度失望，称其充满诉讼与指责，并坦言自己持有的代币已归零，目前生活拮据。颇具讽刺意味的是，其开发的底层开源框架Eliza OS仍在更新，甚至有企业寻求合作，最近一周的主要代码贡献者竟是AI模型Claude。这似乎证明，一个不发币的AI开源项目，生存路径反而更宽广。反思“加密xAI”的现状，单纯的“为AI概念发币”模式已近乎失败。当前仍存活的加密AI项目，如Bittensor、Render，聚焦于用代币协调真实的去中心化算力或渲染资源。另一方面，AI技术正在反向赋能加密领域，如在链上数据分析、交易策略等方面发挥作用。文章最终指出，从业者或许需要调整心态：放弃代币包袱的开发者，在AI世界可能更有价值；而扎实利用AI工具的加密参与者，也能获得比以往更稳健的收益。加密行业常常“快人一步”看到方向，但有时可能搞错了自身该站的位置。

marsbit1小时前

长鑫拒绝苹果压价，价格不低于三星SK海力士，苹果失去了议价权

韩媒报道称，苹果主动寻求向长鑫存储采购DRAM，希望获得比三星、SK海力士更低的价格，但遭到长鑫拒绝。长鑫明确表示，其报价不会低于韩企，甚至可能更高。这背后有几个关键原因。首先，华为、小米等中国本土企业已通过长期合同锁定了长鑫的产能，使其无需迁就苹果的苛刻条件。其次，AI浪潮推动HBM需求爆发，三星、SK海力士将更多产能转向利润更高的HBM，导致通用DRAM供应紧张、价格大幅上涨，苹果成本压力剧增。此外，长鑫的技术已大幅提升，DDR5等产品良率突破90%，与三星的差距显著缩小，不再需要依靠低价竞争。长鑫的拒绝也反映了其对风险的考量。以往依赖苹果供应链的企业（如欧菲光、闻泰科技）曾因订单变动或政治风险遭受重创。长鑫作为受关注的半导体企业，进入果链的政治风险可能大于商业收益。这一事件标志着产业议价权的转移。过去终端品牌主导供应链，如今上游核心产能（尤其是AI相关）成为稀缺资源，供应商话语权增强。中国半导体企业正从“低价替代者”转变为能平等议价、掌握定价权的供应商。苹果试图以长期订单换取低价和优先供货的传统优势正在失效，这背后是中国产业升级和技术实力提升的必然结果。

marsbit1小时前