/
Navigation
Chronicles
Browse all articles
Explore
Semantic exploration
Research
Entity momentum
Nexus
Correlations & relationships
Story Arc
Topic evolution
Drift Map
Semantic trajectory animation
Posts
Analysis & commentary
Pulse API
Tech news intelligence API
Browse
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
Concept Search
Semantic similarity search
High Impact Stories
Top coverage by position
Sentiment Analysis
Positive/negative coverage
Anomaly Detection
Unusual coverage patterns
Analysis
Rivalry Report
Compare two entities head-to-head
Semantic Pivots
Narrative discontinuities
Crisis Response
Event recovery patterns
Connected
Search: /
Command: ⌘K
Embeddings: large
TEXXR

Chronicles

The story behind the story

days · browse · Enter similar · o open

Analysis: Colossus 2, one of the world's largest AI datacenters, uses roughly the same water resources required to supply 2.5 typical In-N-Out Burger locations

It turns out, about twice that of a Burger restaurant. …

SemiAnalysis

Context & Ripple Effects

Colossus 2 had already been framed as an unusually rapid buildout, reaching 200MW in six months while targeting gigawatt-scale capacity in xAI's earlier Colossus 2 construction plan. This analysis adds water use to a story previously dominated by compute scale and chip procurement.

That matters because the project’s expansion has also been tied to plans for roughly 300,000 additional Nvidia chips, making the reported chip buildout a useful indicator of how infrastructure inputs compound as AI campuses grow.

First-order effects

  • The comparison gives Colossus 2’s water footprint a concrete, public benchmark, putting water alongside power and accelerators as an immediate operating constraint for the facility.
  • xAI’s infrastructure narrative now carries a resource-use dimension that local stakeholders and customers can weigh alongside its training capacity.

Second-order effects

  • Other large AI-campus builders will face stronger pressure to explain cooling and water choices in terms that can be compared across facilities, not merely by installed compute or electrical capacity.
  • Infrastructure planning around AI clusters will need to account for water demand together with the power and chip supply requirements already visible in Colossus-scale projects.

Third-order effects

  • If facilities continue to scale toward gigawatt-class AI campuses, water availability and cooling design could become part of the location advantage that determines where frontier training capacity can expand.
  • The broader AI infrastructure race is shifting from procuring chips alone toward managing a bundle of utility inputs, with resource constraints potentially shaping the pace and geography of deployment.

The trend: AI datacenters are becoming utility-scale industrial assets whose competitive limits are set by water, power, cooling, and hardware supply together.

Discussion

  • @thezvi Zvi Mowshowitz on x
    I feel the pain of ‘maybe this new framing will let them notice the math’ so much. We tried comparing queries to hamburgers. How about instead comparing data center water use to an In-N-Out Burger restaurant? Will that get through better? I mean, I assume no, but we continue.
  • @davidsacks David Sacks on x
    Narrative violation: The largest AI data center uses roughly the same amount of water as two burger joints. There are over 200,000 fast food restaurants in the U.S. so this is a tiny amount. [image]
  • @firstadopter Tae Kim on x
    Great piece by @SemiAnalysis_ They ran the numbers on how dumb the medias and certain authors/writers obsession over data center water usage is -> https://newsletter.semianalysis.com/ ... [image]
  • @sriramk Sriram Krishnan on x
    Some remarkable reporting on data center water usage. Going to use this In-N-Out burger comparison now.
  • @druce.ai @druce.ai on bluesky
    SemiAnalysis estimates Colossus 2 uses about 346 million gallons of blue water per year, roughly 2.5 times a typical In-N-Out store's annual use (~147 million gallons).