53 million square feet of data centers have been constructed over the past 20 years
劳登县在过去20年建造了5300万平方英尺的数据中心,平均每年约265万平方英尺。这一规模相当于约244个标准足球场的大小,表明该地区已成为重要的数据中心集群。然而,缺乏与全国其他地区的比较数据,无法确定这一规模是否异常突出。
53 million square feet of data centers have been constructed over the past 20 years
劳登县在过去20年建造了5300万平方英尺的数据中心,平均每年约265万平方英尺。这一规模相当于约244个标准足球场的大小,表明该地区已成为重要的数据中心集群。然而,缺乏与全国其他地区的比较数据,无法确定这一规模是否异常突出。
$130 billion in data center projects blocked by protests so far this year
这一数据点表明,2026年前三个月因抗议而被阻止或延迟的数据中心项目价值高达1300亿美元,占2025年全年记录的1560亿美元的约83%。这一数字反映了数据中心反对运动的显著增长趋势,可能对AI基础设施建设产生重大影响,但需要确认这些数据的统计方法和来源可靠性。
Google will pay SpaceX $920M per month for compute
Google将每月向SpaceX支付9.2亿美元用于计算资源,这一金额极其庞大,年化可达110亿美元。这笔交易表明大型科技公司愿意为计算能力支付高额费用,但也反映出SpaceX在AI基础设施市场的战略定位。然而,如此高额的月度合同是否可持续,以及这是否代表真正的市场认可,仍需观察。这一数字也凸显了AI计算成本的高昂和竞争的激烈程度。
This optimization reduced 'write amplification'—the ratio of data written to storage versus the original request—by 20%. It also provided insights for new compiler optimization strategies that reduced the storage footprint of software by nearly 9%.
除了20%的写入放大减少,AlphaEvolve还通过新的编译器优化策略将软件存储占用减少了近9%。这表明该系统在多个层面优化基础设施的能力,从硬件到软件栈都带来了显著效率提升。
By late 2025, total AI data center power capacity had reached roughly tens of gigawatts, which puts AI's electricity consumption at a scale comparable to the peak electricity demand of the state of New York
AI数据中心总电力容量已达数十吉瓦,相当于纽约州高峰电力需求。这一数据点突显了AI产业对能源的巨大需求,以及由此带来的能源挑战和环境影响。随着AI计算能力继续增长,能源供应将成为制约AI发展的关键因素之一,可能推动行业向更节能的技术方向发展。
up to 5 gigawatts (GW) of capacity for training and deploying Claude
5GW的算力规模是惊人的,相当于一个小型国家的电力消耗。这一数据表明Anthropic正在为AI模型训练和部署投入前所未有的基础设施资源,反映了大语言模型对计算资源需求的指数级增长。这一规模超过了大多数AI公司的基础设施投入,显示出Anthropic在AI基础设施竞争中的野心。
up to 5 gigawatts (GW) of capacity for training and deploying Claude
5GW的算力规模极其庞大,相当于一个小型国家的电力消耗。这一数字表明Anthropic正在为AI模型训练和部署构建前所未有的基础设施,反映了大型语言模型对计算资源的巨大需求。相比其他AI公司的算力规模,这是一个非常激进的扩张计划。
up to 5 gigawatts (GW) of capacity for training and deploying Claude
5GW的算力规模是惊人的,相当于一个小型国家的电力消耗。这个数字表明Anthropic正在为AI模型训练和部署进行大规模基础设施投资,反映了大型语言模型对计算资源的巨大需求。这一规模与OpenAI等竞争对手的算力投入相当,显示AI算力竞赛正在升级。
We need, like, a Manhattan Project to collect this
经济学家呼吁以“曼哈顿计划”的规模来收集各行业价格弹性数据,凸显了当前AI经济研究的底层基础设施极度匮乏。没有跨经济体的系统性微观数据支撑,任何关于AI就业前景的预测都是盲人摸象,政策制定更是无从谈起。
During each call, Stewart said, Amazon officials have not been helpful."They wanted to do background checks on all my firefighters; I wouldn't let them," he said. "And we've struggled to gain access to emergencies. They'll stop us at the gate, and our medic units have been delayed. They're denying us access to patients.
"How AI Datacenters Eat the World" from High Yield on YouTube. 30-Aug-2025
HighYield x SemiAnalysis deep-dive into AI Datacenters, Gigawatt Megaclusters and the Hyperscaler race to AGI. How AI Datacenters Eat the World.
Recently, OpenAI has shared something. In a blog post, CEO Sam Altman said that the average query uses about 0.34 watt hours of energy.
From the 10-Jun-2025 blog post:
People are often curious about how much energy a ChatGPT query uses; the average query uses about 0.34 watt-hours, about what an oven would use in a little over one second, or a high-efficiency lightbulb would use in a couple of minutes. It also uses about 0.000085 gallons of water; roughly one fifteenth of a teaspoon.
nebraska case study of data sharing for court-involved youth
Synergies for Europe's Research Infrastructures in the Social Sciences (SERISS)
How can we build an open community-led commons of grants data?
No, United Way isn’t sitting still—it recently teamed up with Salesforce.org to roll out a new app called Philanthropy Cloud. For now, however, upstart Benevity rules the online workplace giving space.
Malamud’s General Index
Weapons of Affect: The Imperative for Transdisciplinary Information Systems Design
Webpages for hundreds of hospitals require users to click through to find prices, undermining federal transparency rule, Journal analysis shows
Ali, A. (2020, August 28). Visualizing the Social Media Universe in 2020. Visual Capitalist. https://www.visualcapitalist.com/visualizing-the-social-media-universe-in-2020/
Welcome! You are invited to join a webinar: Supporting Open Science Data Curation, Preservation, and Access by Libraries. After registering, you will receive a confirmation email about joining the webinar. (n.d.). Zoom Video. Retrieved June 28, 2020, from https://zoom.us/webinar/register/2615905946283/WN_W6dYUXQFTqGQjGAZPRB74w
Center for Scientific Workshops in All Disciplines—Lorentz-eScience Competition. (n.d.). Retrieved April 16, 2020, from https://www.lorentzcenter.nl/lorentz-escience-competition.html
Draft NIH Policy for Data Management and Sharing
pretty great intro to knowledge graphs
Instead of encouraging more “data-sharing”, the focus should be the cultivation of “data infrastructure”,¹⁴ maintained for the public good by institutions with clear responsibilities and lines of accountability.
We showhow the rise of large datasets, in conjunction with arising interest in data as scholarly output, contributesto the advent of data sharing platforms in a field trad-itionally organized by infrastructures.
What does this paper mean by infrastructures? Perhaps this is a reference to the traditional scholarly journals and monographs.
This approach is called change data capture, which I wrote about recently (and implemented on PostgreSQL). As long as you’re only writing to a single database (not doing dual writes), and getting the log of writes from the database (in the order in which they were committed to the DB), then this approach works just as well as making your writes to the log directly.
Interesting section on applying log-orientated approaches to existing systems.