{
  "schemaVersion": 1,
  "id": "atlas-lesson-regional-siting-access",
  "title": "Siting, regional infrastructure and access to compute",
  "category": "learning",
  "format": "Atlas lesson",
  "description": "Connects location choices to service paths, resources and who can actually use capacity.",
  "canonicalUrl": "https://semiaifoundry.com/learn/atlas/regional-siting-access/",
  "publishedDate": null,
  "revisionDate": "2026-10-07",
  "author": null,
  "topic": "Global AI infrastructure",
  "series": null,
  "license": null,
  "rightsUrl": "https://semiaifoundry.com/ai-atlas/licenses.html",
  "version": "1.3.0",
  "status": "Published educational lesson",
  "evidenceScope": "No jurisdiction-specific legal advice or exhaustive regional census is provided. Public project and procurement statements support dated examples only.",
  "contentCoverage": "published-lesson",
  "artifactUrl": "https://semiaifoundry.com/ai-atlas/lesson-data/regional-siting-access-797dbbfb173b8938.json",
  "interactiveUrl": "https://semiaifoundry.com/ai-atlas/#regional-siting-access",
  "related": [],
  "supersedes": [],
  "supersededBy": [],
  "contentHash": "1a3129d740049b56694991840e8c05f3aee9e88a5cf4791f751414d6b9a1fe96",
  "contentLength": 5348,
  "jsonUrl": "https://semiaifoundry.com/agents/publications/atlas-lesson-regional-siting-access.json",
  "markdownUrl": "https://semiaifoundry.com/agents/publications/atlas-lesson-regional-siting-access.md",
  "sections": [
    {
      "id": "section-1",
      "title": "Siting, regional infrastructure and access to compute",
      "start": 41,
      "end": 441
    },
    {
      "id": "section-2",
      "title": "The mechanism",
      "start": 441,
      "end": 1577
    },
    {
      "id": "section-3",
      "title": "Science and assumptions",
      "start": 1577,
      "end": 1605
    },
    {
      "id": "section-4",
      "title": "Conditional site score",
      "start": 1605,
      "end": 2400
    },
    {
      "id": "section-5",
      "title": "Engineering example",
      "start": 2400,
      "end": 2630
    },
    {
      "id": "section-6",
      "title": "Distinctions and tradeoffs",
      "start": 2630,
      "end": 3088
    },
    {
      "id": "section-7",
      "title": "Specifications and interfaces",
      "start": 3088,
      "end": 3368
    },
    {
      "id": "section-8",
      "title": "Interfaces",
      "start": 3368,
      "end": 3556
    },
    {
      "id": "section-9",
      "title": "Check your understanding",
      "start": 3556,
      "end": 3585
    },
    {
      "id": "section-10",
      "title": "Which question distinguishes accessible capacity from physical capacity?",
      "start": 3585,
      "end": 4001
    },
    {
      "id": "section-11",
      "title": "Evidence and scope",
      "start": 4001,
      "end": 4279
    },
    {
      "id": "section-12",
      "title": "Sources",
      "start": 4279,
      "end": 4822
    },
    {
      "id": "section-13",
      "title": "Continue exploring",
      "start": 4822,
      "end": 4845
    },
    {
      "id": "section-14",
      "title": "Foundations",
      "start": 4845,
      "end": 5074
    },
    {
      "id": "section-15",
      "title": "Related lessons",
      "start": 5074,
      "end": 5348
    }
  ],
  "citations": [
    {
      "title": "Explore in Atlas",
      "url": "https://semiaifoundry.com/ai-atlas/#regional-siting-access"
    },
    {
      "title": "Google AI Hub construction milestone in India",
      "url": "https://blog.google/intl/en-in/google-ai-hub-scaling-indias-ai-infrastructure-and-community-impact/"
    },
    {
      "title": "The EuroHPC Joint Undertaking launches the AI Gigafactories Call",
      "url": "https://www.eurohpc-ju.europa.eu/eurohpc-joint-undertaking-launches-ai-gigafactories-call-2026-07-30_en"
    },
    {
      "title": "Published lesson data",
      "url": "https://semiaifoundry.com/ai-atlas/lesson-data/regional-siting-access-797dbbfb173b8938.json"
    },
    {
      "title": "Atlas releases and licensing",
      "url": "https://semiaifoundry.com/ai-atlas/licenses.html"
    },
    {
      "title": "Facility phases and geographical evidence",
      "url": "https://semiaifoundry.com/learn/atlas/capacity-phase-geography/"
    },
    {
      "title": "Cluster scheduler and resource allocation",
      "url": "https://semiaifoundry.com/learn/atlas/scheduler/"
    },
    {
      "title": "Grid interconnection and deliverable electrical capacity",
      "url": "https://semiaifoundry.com/learn/atlas/grid-interconnection-readiness/"
    },
    {
      "title": "Job admission, topology and resource fragmentation",
      "url": "https://semiaifoundry.com/learn/atlas/topology-job-admission/"
    }
  ],
  "recordHash": "50448a99fc08ca02580f6b99fde7885bda85e817b8d27b305e2fc501079d326b",
  "contentFormat": "text/markdown",
  "content": "Global AI infrastructure · Atlas v1.3.0\n\n# Siting, regional infrastructure and access to compute\n\nConnects location choices to service paths, resources and who can actually use capacity.\n\n[Explore in Atlas](https://semiaifoundry.com/ai-atlas/#regional-siting-access)Facility phases to region and world; geography schematic\n\nLearning objectiveExplain siting and access tradeoffs without equating investment, geography and available service.\n\n## The mechanism\n\nData-center siting combines electricity readiness, fiber routes, latency, climate, water, staffing and the needs of its users. A project can improve connectivity or electrical infrastructure while remaining under construction. Google’s India hub announcement explicitly links the site with surrounding infrastructure work, while EuroHPC’s gigafactory call describes a public access and procurement mechanism. These are different evidence categories from private installed hardware. Owner, operator and access model are separate attributes. Announced investment does not equal completed equipment, and nearby compute is not automatically available to every local user.\n\n1.  Define the users, data paths and service objectives that location must support.\n2.  Compare actual electrical, network and cooling readiness for the same time horizon.\n3.  Identify physical operators and the mechanism through which users receive compute access.\n4.  Model resource and latency tradeoffs with explicitly chosen scenario weights.\n5.  Keep construction announcements, funding commitments and operational service evidence separate.\n\n## Science and assumptions\n\n### Conditional site score\n\n$$\n\\mathrm{Score}=\\sum_i w_i z_i\n$$\n\nThe scenario score sums normalized criterion values multiplied by declared preference weights.\n\nThis is a preference-dependent decision aid, not an objective scientific ranking. Weight normalization is a chosen convention, not implied by the sum.\n\nSymbols and units\n\nScore\n\nConditional scenario preference score\n\ndimensionless\n\nwᵢ\n\nDeclared preference weight for criterion i\n\ndimensionless\n\nzᵢ\n\nNormalized metric value under the stated transformation\n\ndimensionless\n\ni\n\nCriterion index\n\ninteger index\n\nzi is a normalized scenario metric and wi a declared preference weight; the score is dimensionless.\n\nAssumptions A teaching decision aid, not an objective scientific ranking; normalization and uncertain inputs must be visible.\n\n## Engineering example\n\nSynthetic choice: Site A offers lower network delay but opens a year later; Site B opens sooner with a longer path. A batch-training user and an interactive-serving user can reasonably choose differently.\n\n## Distinctions and tradeoffs\n\nInvestment dollars are not operating megawatts, and public access programs do not imply every listed facility is already commissioned.\n\nProximity can reduce network delay while a different region may have earlier power availability; access rules and reservations can dominate physical nearness.\n\nUseful compute capacity depends on admission and data paths, so geographical expansion matters through more than building counts.\n\n## Specifications and interfaces\n\nAccess states\n\nOwned, reserved, allocated and available\n\nPhysical capacity and user access are separate ledgers.\n\nLocation objective\n\nLatency, readiness, resilience and resources\n\nWeights belong to an explicit scenario, not a universal ranking.\n\n### Interfaces\n\nWorkload and regional requirements Input\n\nUser locations, resource limits and access needs\n\nSiting/access decision Output\n\nA bounded choice with evidence-backed tradeoffs\n\n## Check your understanding\n\n### Which question distinguishes accessible capacity from physical capacity?\n\n1.  How many countries appear in the press release?\n2.  Who can obtain an allocation, under what conditions and date?\n3.  How large is the site’s sign?\n\nAnswer and reasoning\n\nWho can obtain an allocation, under what conditions and date?\n\nAccess depends on reservations, procurement and admission policy beyond the presence of equipment.\n\n## Evidence and scope\n\nNo jurisdiction-specific legal advice or exhaustive regional census is provided. Public project and procurement statements support dated examples only.\n\nBrowse the numerical library; no current calculator directly models this lesson’s central mechanism.\n\n## Sources\n\n-   [Google AI Hub construction milestone in India](https://blog.google/intl/en-in/google-ai-hub-scaling-indias-ai-infrastructure-and-community-impact/)\n-   [The EuroHPC Joint Undertaking launches the AI Gigafactories Call](https://www.eurohpc-ju.europa.eu/eurohpc-joint-undertaking-launches-ai-gigafactories-call-2026-07-30_en)\n\n[Published lesson data](https://semiaifoundry.com/ai-atlas/lesson-data/regional-siting-access-797dbbfb173b8938.json) · [Atlas releases and licensing](https://semiaifoundry.com/ai-atlas/licenses.html)\n\n## Continue exploring\n\n### Foundations\n\n-   [Facility phases and geographical evidence](https://semiaifoundry.com/learn/atlas/capacity-phase-geography/)\n-   [Cluster scheduler and resource allocation](https://semiaifoundry.com/learn/atlas/scheduler/)\n\n### Related lessons\n\n-   [Grid interconnection and deliverable electrical capacity](https://semiaifoundry.com/learn/atlas/grid-interconnection-readiness/)\n-   [Job admission, topology and resource fragmentation](https://semiaifoundry.com/learn/atlas/topology-job-admission/)"
}