LIVE
QUEUE 438.0 GWDEMAND 60.4 GWAPPROVED 9,042 MWOBSERVED 3,883 MW|The New York Times Meta in Talks to Lease Computing Power to Anthropic in Potential $10 Billion Deal - The New York Times|Yahoo Huge AI data center rejected in Florida. Why are they so controversial? - Yahoo|Newsday Brookhaven Town OKs 18-month AI data center moratorium - Newsday|CNBC Anthropic in early talks with Meta to acquire compute power - CNBC|Austin American-States 'Prefer an outright ban': Small Texas town weighs data center restrictions amid industry growth - Austin American-Statesman|NextBigFuture.com SpaceXAI Provides Near IMMEDIATE Gigawatt Scale AI Data Center Capacity - NextBigFuture.com|DCD data center news Coalition of environmental action groups petition FCC to halt orbital data center applications|Utility Dive AI data center growth could force US utilities to rethink generation plans, BofA says - Utility Dive|Data Center Dynamics Duos Edge AI to provide 10MW of capacity for five years to “investment-grade hyperscaler” - Data Center Dynamics|Yahoo! Finance Canada Pennsylvania families sold 1,700 acres for $586 million as AI data center land rush mints new millionaires - Yahoo! Finance Canada|Lexington Herald Leade Kentucky powered America once. It can help power the AI era too | Opinion - Lexington Herald Leader|moneywise.com Pennsylvania pig farmers sell 89-acre farm for $22 million as AI data center boom mints dozens of new millionaires - moneywise.com|The Motley Fool 3 Utility Stocks Built for the Coming AI Power Crunch - The Motley Fool|MyHighPlains.com Crusoe, Lancium announce partnership for 1-gigawatt AI data center campus in Childress - MyHighPlains.com|DCD data center news Elon Musk quietly acquires mobile gas generation firm APR Energy - report|The Globe and Mail 3 Utility Stocks Built for the Coming AI Power Crunch - The Globe and Mail|QUEUE 438.0 GWDEMAND 60.4 GWAPPROVED 9,042 MWOBSERVED 3,883 MW|The New York Times Meta in Talks to Lease Computing Power to Anthropic in Potential $10 Billion Deal - The New York Times|Yahoo Huge AI data center rejected in Florida. Why are they so controversial? - Yahoo|Newsday Brookhaven Town OKs 18-month AI data center moratorium - Newsday|CNBC Anthropic in early talks with Meta to acquire compute power - CNBC|Austin American-States 'Prefer an outright ban': Small Texas town weighs data center restrictions amid industry growth - Austin American-Statesman|NextBigFuture.com SpaceXAI Provides Near IMMEDIATE Gigawatt Scale AI Data Center Capacity - NextBigFuture.com|DCD data center news Coalition of environmental action groups petition FCC to halt orbital data center applications|Utility Dive AI data center growth could force US utilities to rethink generation plans, BofA says - Utility Dive|Data Center Dynamics Duos Edge AI to provide 10MW of capacity for five years to “investment-grade hyperscaler” - Data Center Dynamics|Yahoo! Finance Canada Pennsylvania families sold 1,700 acres for $586 million as AI data center land rush mints new millionaires - Yahoo! Finance Canada|Lexington Herald Leade Kentucky powered America once. It can help power the AI era too | Opinion - Lexington Herald Leader|moneywise.com Pennsylvania pig farmers sell 89-acre farm for $22 million as AI data center boom mints dozens of new millionaires - moneywise.com|The Motley Fool 3 Utility Stocks Built for the Coming AI Power Crunch - The Motley Fool|MyHighPlains.com Crusoe, Lancium announce partnership for 1-gigawatt AI data center campus in Childress - MyHighPlains.com|DCD data center news Elon Musk quietly acquires mobile gas generation firm APR Energy - report|The Globe and Mail 3 Utility Stocks Built for the Coming AI Power Crunch - The Globe and Mail
Confirmation email sentOpen the link from Data Center Watch to finish subscribing. Check spam if needed.
Confirmation email sentOpen the link from Data Center Watch to finish subscribing. Check spam if needed.
Confirmation email sentOpen the job-alerts link from Data Center Watch to finish. Check spam if needed.
Data Center

WATCH

power · compute · buying · dependencies
ERCOT CAISO PJM MISO NYISO ISO-NE SPP
Global data-center intelligence

Every megawatt
has a fingerprint.

Track facilities, current load, future requests, power-capacity gaps, component BOMs, buying intent, producer constraints, and the primary → secondary → tertiary dependencies that turn a greenfield site into an operating compute campus. Every number links to evidence.

ERCOT deep dive · evidence store · 2026-07-17
Get the signal →
Initializing orbital network map…
15,512 data centers
PAN
drag · scroll
+/− · F full
Loading map detail…
15,512data centers
7,789localities
32fiber arcs
179countries
15,512 sites signals 179 countries
lat — · lng — · alt —
Global Data Center Intelligence Summary — zsh — 80×24
$ ./disambiguate --region=ERCOT --style=casual
Here's where things stand in ERCOT today: there's 438.0 GW worth of new data centers trying to plug into the grid, but only 3,883 MW of that is actually pulling power right now. That leaves a 434.1 GW gap between what's been requested and what's real — some of that is projects still ramping up, some may never actually show up. Overall grid demand right now is 60.4 GW, which is about 35% of everything the region can generate. Most of that generating power (174.9 GW total) is coming from natural gas. On the AI-compute front, the biggest system we're tracking is El Capitan, clocking in around 1,742.0 petaflops. Zoom out and there are 27,403 data centers catalogued worldwide right now, from ordinary colo buildings to a handful of facilities that aren't even on Earth. Get this delivered daily → https://datacentr.net/#digest
Direct / operatingmetered load, fleet, realized capacity
Buyer / forwardRFPs, construction spend, queue requests
Producer / upstreamrevenue, trade, certification, backlog
Tertiary / constraintwater, permits, climate, community risk

Signal board

Every headline datapoint is a card. Click any card to jump to the underlying detail.

ERCOT large-load queue
future
438.0 GW
+94% vs. 2025
New submissions to 2036
future
140.0 GW
reported submissions still being processed
Queue attributed to data centers
future
90%
editorial feed signal · not a forecast
Approved to energize
process
9,042 MW
ERCOT · cumulative
Observed operational peak
live
3,883 MW
large loads actually drawing power
Queue not yet observed
gap
434.1 GW
gross backlog proxy · not a cancellation estimate
Power gap proxy
gap
263.1 GW
queue − fleet · 2.5× current nameplate
ERCOT demand, live
live
60.4 GW
EIA-930 · 2026-07-17T05Z · 35% of nameplate
Energy delivered
live
66,938 GWh
ERCOT · last 42 days
Operating generation fleet
live
174.9 GW
1 region · EIA-860/860M
Fuel + firming base
live
67.2 GW
Natural Gas · 38% · battery 15.9 GW
Flagship compute
live
1,742.0 PF
El Capitan · 29.6 MW · queue ≈ 14,807×
Compute efficiency
live
58.9 GF/W
134× since 2008
Data-center build cost
process
$11–11.3/W
full build / shell-and-core benchmarks
Cooling efficiency · best fleet
process
1.09 PUE
Google (global fleet-wid · broad avg 1.54
Next Batch Zero deadline
process
7 days
2026-07-24 · Batch Zero Load Information Form due t
Extreme-environment compute
coverage
38 records
21 deployed · 1 operating · 7 planned · 5 study · atmosphere 5, floating 6, lunar 1, ocean 7, orbit 8, underground 11
Data center jobs
process
171 open roles
employer ATS boards · Texas & remote
Worldwide DC facilities
coverage
27,403
15,512 geocoded · 179 countries · open sources
Facilities with disclosed MW
live
583 sites
Σ disclosed 240.5 GW · never estimated for gaps
Component buying profiles
coverage
24 source-locked
CPU · GPU · HBM · DIMMs · network · power · cooling
Buyer + producer feeds
future
39 added
RFPs · construction · revenue · trade · certification
Worldwide source registry
coverage
105 sources
buyer intent · producer supply · power · water · silicon
Coverage footprint
coverage
13 dependency layers
7/7 regions live · primary → tertiary

Worldwide facility registry 27,403 sites · open sources only · every row sourced

Worldwide facility layer assembled from open sources plus the full ATLAS directory (unrestricted use granted to Data Center Watch by the ATLAS IP team, 2026-07), OpenStreetMap (ODbL), PeeringDB interconnection campuses, FracTracker US operating/proposed inventory, and Epoch AI large AI campuses (CC BY 4.0). ATLAS coordinates are city-level centroids (labelled as such), not rooftop GPS. DataCenterMap and Baxtel remain catalogued only (commercial no-scrape). MW and H100-eq appear only when a source states them. Gaps stay empty. The globe plots every geocoded row; fiber arcs remain illustrative corridors.

Facility count by country label · incomplete where source omits country

Provenance mix · dual-count possible pre-dedup across sources

FacilityOperatorCountryCity / regionMWH100-eqStatusSource
Project JadeCrusoeUnited StatesCheyenne10.0 GWsuspendedFracTracker Alliance National Da
SoftBank Data Center/PORTS Technology CampusSoftbank/Open AI (unconfirmed)United StatesPiketon9,200 MWapproved/permitted/under conFracTracker Alliance National Da
Stratos Project, eastern parcelsUnited StatesSnowville9,000 MWapproved/permitted/under conFracTracker Alliance National Da
New Era Energy & Digital Data CenterUnited StatesLea County7,000 MWproposedFracTracker Alliance National Da
GW Ranch (Pacifico Energy)United StatesPecos County5,000 MWproposedFracTracker Alliance National Da
Homer City Energy CampusHomer City Redevelopment (with Kiewit as contractor/partner)United StatesHomer City4,500 MWapproved/permitted/under conFracTracker Alliance National Da
Shippingport Bruce Mansfield Data Center/Project PhoenixAligned Data CentersUnited StatesShippingport3,600 MWapproved/permitted/under conFracTracker Alliance National Da
Pittsylvania County Data CenterBalico LLCUnited StatesChatham3,500 MWcancelledFracTracker Alliance National Da
MSB Global Services Matrix Data CenterUnited StatesSulphur Springs3,000 MWproposedFracTracker Alliance National Da
Stak Energy Data CenterStakUnited StatesNorth Slope Borough3,000 MWproposedFracTracker Alliance National Da
Vermaland Data CenterVermaland LLCUnited StatesRed Rock3,000 MWproposedFracTracker Alliance National Da
Hobart Data CenterAmazonUnited StatesHobart2,400 MWapproved/permitted/under conFracTracker Alliance National Da
Tract Technology CampusVALCO HANOVER COUNTYUnited StatesAshland2,400 MWproposedFracTracker Alliance National Da
Adams Fork Harless Data CenterTransGas Development, LLCUnited StatesHolden2,398 MWproposedFracTracker Alliance National Da
Adams Fork Wharncliffe Data CenterTransGas Development, LLCUnited StatesWharncliffe2,398 MWproposedFracTracker Alliance National Da

Showing top 80 by disclosed power / connectivity density of 27,403. Full machine-readable export: JSON · CSV · 15,512 geocoded · 583 with MW · 179 countries.

Buildout dependency map the chain behind every new data center

Coverage model: what we already observe, what should be ingested next, and where a power story becomes a land, water, equipment, labor, capital, or community story. Green dots are tracked; violet dots are catalogued for ingestion.

News monitor daily crawl · headline links only · source first

The ticker is fed by an automated RSS crawl across grid, energy, data-center, HPC, and large-load sources. Headlines link to the original publisher; Data Center Watch does not copy article bodies.

Grid pulse hourly demand + daily peaks · EIA-930

Actual system demand — not interconnection requests. Crosshair cursors, ±1σ band, mean reference, scientific axes. Gaps stay empty — nothing is interpolated.

μ / σ / range from observed hours only · brush + scroll to zoom · EIA-930

Daily peak by balancing authority · fixed categorical colors · nulls not filled

Interconnection queues latest verified totals · ERCOT history

Step series · queue jumps only at published decks · no interpolation between months

Approved vs observed · realization % = observed ÷ approved when both exist

RegionTotal queueAs of
ERCOT438.0 GW2026-03-31

Generation capacity actual in-service fleet · EIA-860/860M

Nameplate capacity by fuel — what is built, not what is requested. Complements the queue charts above.

Nightingale rose · nameplate MW share · EIA-860/860M

Lollipop fleet map · operating nameplate by region

RegionMWAs of
ERCOT174.9 GW2026-04
ERCOT fuelMWShare
Natural Gas67.2 GW38%
Wind40.0 GW23%
Solar31.6 GW18%
Battery Storage15.9 GW9%
Coal13.4 GW8%
Nuclear5,139 MW3%
Oil/Diesel798 MW0%
Hydro549 MW0%
Other/Biomass374 MW0%

Industry benchmarks every figure independently verified against its primary source

Lower is better · reference line at industry ~1.5 PUE · source-linked table below

CapEx per watt

Entity$/WAs ofSource
1x1 combined-cycle gas turbine plant (generation side, not the data center building)$0.92/W2023GridLab / Energy Futures Group / Halcyon, citing EIA (Sargent & Lundy cost study)
Traditional prime markets, full build (shell + M&E + cooling + fit-out)$11.00/W2025-2026CBRE Investment Management, citing Turner & Townsend Data Center Cost Index
Global average (shell-and-core construction)$11.30/W2026 (forecast)JLL

Cooling — PUE table

EntityPUEAs ofSource
Google (global fleet-wide)1.092025Google (datacenters.google)
AWS (global fleet-wide)1.142025Amazon.com Inc. (Amazon Sustainability)
US national average (all data center types)1.402023Lawrence Berkeley National Laboratory — 2024 United States Data Center Energy Usage Report (LBNL-2001637)
Global industry average (Uptime Institute Global Data Center Survey 2025, n=681)1.542025Uptime Institute (Uptime Intelligence)

DC share of annual grid electricity

EntityShareAs ofSource
United States — all data centers (649 TWh reference case)11.8% of total U.S. electricity use2030 (reference forecast)Lawrence Berkeley National Laboratory — United States Data Center Energy Usage Report: 2025 Update
Ireland — all grid-connected data centers (7.663 TWh metered)23.0% of total metered electricity consumption2025Central Statistics Office Ireland — Data Centres Metered Electricity Consumption 2025
United States — all data centers (176 TWh estimated consumption)4.4% of total U.S. electricity use2023Lawrence Berkeley National Laboratory — 2024 United States Data Center Energy Usage Report

Buildout & dependency chain

MetricEntityValueAs ofSource
Capacity auction clearing pricePJM RTO (2027/2028 Base Residual Auction; up from $329.17 in 2026/27)333.44 USD/MW-day2027/2028 delivery year (BRA report published 2025)
Contracted large-load capacityIreland — data centers + other new-technology loads, transmission contracts2,000 MW2025-08
Data-center capacityNew Zealand — data-center capacity (median scenario)400 MW2035 (forecast)
Data-center energy shareIreland — data centers + new-technology loads (median scenario)30.0% of electrical energy requirements2032 (forecast)
Projected data-center load growthPJM — projected data-center load growth (up to)30.0 GW2030 (forecast)

Infrastructure scoper CPU · GPU · HBM · DIMMs · power train · cooling loop

Translate a compute-power target into a first-pass equipment envelope using primary manufacturer specifications. Reported facts and derived arithmetic are labelled separately; unknowns stay unknown.

AI factory sizing console

Fast screening for owner briefs, buying plans, and early RFP scope — not electrical or mechanical design.

source lockedtransparent mathno gap filling
Deployment unitsCeiling of target ÷ published unit power
AcceleratorsPhysical GPU / OAM count
Host CPUsOnly when the profile states them
Accelerator memoryDecimal PB of HBM
Facility demandCompute target × PUE
Utility serviceFacility MW ÷ power factor
Backup generator unitsInstalled screen ÷ selected genset MW
Cooling-loop flowWater equivalent at selected ΔT; not consumption
Annual site energyAt full target, 8,760 hours
Loading the source profile…

Screening formulae · units = ceil(target MW × 1,000 ÷ unit kW) · facility MW = target × PUE · utility MVA = facility MW ÷ PF · loop gpm = target MW × 3,412,142 ÷ (500 × ΔT). N+1 is a user-selected 1.25× allowance, not an engineered topology. Water consumption is never inferred from loop flow.

Component BOM library manufacturer facts, explicit arithmetic, honest unknowns

NVIDIA · full rackreported + derived

GB300 NVL72

72GPU36CPU20.7 TBaccelerator memory142.0 kWpower basis

Cooling · Direct liquid cooling at the rack/compute-tray level

Memory / DIMMs · Grace uses soldered LPDDR5; 1 TB aggregate CPU main memory per compute tray, not field-replaceable DIMMs.

Network · Derived rack total: 72 x ConnectX-8 HCAs and 18 x BlueField-3 DPUs from the per-tray specification.

142 kW is a published maximum for the full rack, not a measured workload average. Memory uses decimal GB. Storage count excludes management nodes and external storage. Installed PSU nameplate capacity must not be treated as rack draw.

NVIDIA Enterprise Reference Architecture: GB300 NVL72 Components ↗
NVIDIA · compute traymanufacturer reported

GB300 NVL72 compute tray

4GPU2CPU1.2 TBaccelerator memorynot disclosedpower basis

Cooling · Direct liquid cooling

Memory / DIMMs · 1 TB aggregate LPDDR5 CPU main memory across two Grace CPUs; no replaceable DIMM population is specified.

Network · 4 x ConnectX-8 HCAs and 1 x BlueField-3 DPU per compute tray.

A tray is a component of the NVL72 system rather than a standalone server bill of materials. The five-device count is 1 M.2 OS device plus 4 E1.S NVMe cache devices and does not express capacity.

NVIDIA Enterprise Reference Architecture: GB300 NVL72 Components ↗
NVIDIA reference design · control nodereported + derived

Recommended x86 control node for GB300 NVL72 AI Factory

0GPU2CPU0 GBaccelerator memorynot disclosedpower basis

Cooling · Not prescribed; depends on the selected OEM control node

Memory / DIMMs · 8 x 64 GB DDR5 DIMMs = 512 GB.

Network · 4 x ConnectX-7 200 Gb/s adapters.

The architecture recommends 12 management/control nodes, but this profile describes one node. CPU wording follows the reference configuration; it is not an endorsement of a particular CPU vendor or SKU.

NVIDIA Enterprise Reference Architecture: GB300 NVL72 Components ↗
AMD · oam acceleratormanufacturer reported

Instinct MI355X accelerator

1GPU0CPU288 GBaccelerator memory1.4 kWpower basis

Cooling · Platform implementation may use passive or active cooling

Memory / DIMMs · 288 GB HBM3E accelerator memory; no host DIMM requirement is defined by the accelerator page.

Network · PCIe 5.0 x16 host interface; fabric and NIC design are platform-specific.

Do not use 1.4 kW as server or rack input power. AMD also reports 8 TB/s peak memory bandwidth and 185 billion transistors; neither is a facility-load metric.

AMD Instinct MI355X accelerator specifications ↗
AMD · eight accelerator screening profilereported + derived

8 x Instinct MI355X accelerator platform

8GPUCPU2.3 TBaccelerator memory11.2 kWpower basis

Cooling · OEM/platform-specific; both passive and active accelerator implementations are supported

Memory / DIMMs · 2,304 GB aggregate HBM3E is derived; host CPU memory and DIMM population are unspecified.

Network · Host, scale-up fabric, NIC, DPU, and switch requirements are not included.

This is transparent screening arithmetic, not an AMD server BOM. It omits CPUs, DIMMs, storage, networking, cooling auxiliaries, PSUs, conversion losses, and redundancy; it cannot be compared directly with the complete 142 kW NVL72 rack rating.

AMD Instinct MI355X accelerator specifications ↗
NVIDIA · full rackreported + derived

NVIDIA DGX GB200 NVL72 (rack system)

72GPU36CPU13.4 TBaccelerator memory120.0 kWpower basis

Cooling · Hybrid cooling: GPUs and CPUs are direct liquid-cooled via cold plates fed by rack manifolds; networking and storage devices are air-cooled (per the same user guide).

Memory / DIMMs · Not stated in the reviewed NVIDIA GB200 documentation at rack-spec level (no CPU DIMM/memory-channel table published); accelerator memory is separately stated as 'Up to 13.4 TB HBM3e | 576 TB/s' on NVIDIA's DGX GB200 product page.

Network · Per compute tray (18 trays/rack): 4x NVIDIA ConnectX-7 single-port 400G OSFP NIC for cross-rack compute networking; 2x NVIDIA BlueField-3 DPU, dual-port 400G InfiniBand/Ethernet, for storage/management. Out-of-band: 1x 1GbE RJ45 from the compute tray BMC plus 2x 1GbE RJ45 from the BlueField-3 BMC interface. Rack also includes 9 NVLink switch trays providing the 72-GPU NVLink domain.

Covers only the DGX GB200 NVL72 rack itself (compute trays, NVLink switch trays, power shelves). Does not include external InfiniBand/Ethernet leaf-spine switches, external high-performance storage appliances, management nodes, or facility cooling (CDU/pump) overhead -- these are shown as separate racks in NVIDIA's own SuperPOD Scalable Unit diagrams and are folded into the separate, not-directly-comparable 1.2MW per-SU TDP figure.

NVIDIA DGX GB Rack Scale Systems User Guide -- Hardware (power, storage, cooling, networking); NVIDIA DGX GB200 product page (GPU memory figure); NVIDIA DGX SuperPOD Reference Architecture Featuring NVIDIA DGX GB200, RA-11338-001 (compute tray CPU/GPU composition) ↗
NVIDIA · full serverreported + derived

NVIDIA DGX B200

8GPU2CPU1.4 TBaccelerator memory14.3 kWpower basis

Cooling · Air-cooled, 10 RU rack-mount chassis (per the datasheet's Rack Units spec).

Memory / DIMMs · 2TB system memory, configurable to 4TB, per the official datasheet ('System Memory: 2TB, configurable to 4TB'); specific DIMM count/type is not published in the datasheet.

Network · 4x OSFP ports serving 8x single-port NVIDIA ConnectX-7 VPI (up to 400Gb/s NVIDIA InfiniBand/Ethernet each); 2x dual-port QSFP112 NVIDIA BlueField-3 DPU (up to 400Gb/s InfiniBand/Ethernet); management network: 10Gb/s onboard NIC with RJ45, 100Gb/s dual-port Ethernet NIC, and host BMC with RJ45.

Power figure covers the DGX B200 server chassis only -- it does not include top-of-rack network switches, external storage, or facility-level PUE overhead. NVIDIA's DGX SuperPOD B200 guidance elsewhere discusses deploying 2 systems per 42U rack, which would roughly double this per-system figure, but no NVIDIA-published dual-system rack total was located for this candidate.

NVIDIA DGX B200 Datasheet ↗
NVIDIA · eight accelerator screening profilereported + derived

NVIDIA HGX B200 (8-GPU baseboard)

8GPUCPU1.4 TBaccelerator memory8.0 kWpower basis

Cooling · Not specified for the baseboard alone. The datasheet states the baseboard 'can be paired with advanced air or liquid cooling solutions and additional server components' -- cooling design is left to the OEM system integrator.

Memory / DIMMs · Not applicable / not stated -- the baseboard itself carries no system DIMMs; host CPU memory is provided by the OEM server board, which is out of scope for this NVIDIA baseboard datasheet.

Network · Not stated in this document. The baseboard integrates fifth-generation NVLink and fifth-generation NVSwitch for GPU-to-GPU interconnect, with 14.4 TB/s total NVLink bandwidth. Host-facing networking (e.g., ConnectX/BlueField NICs) is determined by the OEM system and is not part of the baseboard reference.

Not suitable for facility-level PUE math -- this is a GPU-only TBP sum for a bare 8-GPU OEM baseboard, not a measured or vendor-stated system power draw. Real HGX B200-based servers from OEMs publish their own higher system-level power specs (including CPUs, DIMMs, storage, fans, PSU overhead) that were out of scope for this NVIDIA-primary-source-only candidate.

PCF Summary for NVIDIA HGX B200 | Datasheet ↗
AMD · eight accelerator screening profilereported + derived

AMD Instinct MI300X Platform (Universal Baseboard, 8x OAM)

8GPUCPU1.5 TBaccelerator memory6.0 kWpower basis

Cooling · Not stated in the platform data sheet — no air/liquid designation is given for the 8x OAM UBB baseboard itself.

Memory / DIMMs · Not specified — this data sheet documents the 8x OAM GPU baseboard only; no host CPU/DIMM memory configuration is given.

Network · 8x PCIe Gen 5 x16 (128 GB/s) host I/O, one per GPU; intra-baseboard AMD Infinity Fabric mesh at 7x 128 GB/s per GPU, 896 GB/s peak aggregate bidirectional ring bandwidth.

This is a GPU-baseboard-only figure (8x OAM accelerators on a UBB 2.0), not a full-server or full-rack power draw. It excludes host CPU(s), system memory, storage, networking, fans, PSUs, and AC/DC conversion losses, so it is not suitable for facility-level PUE math on its own — only for screening accelerator-silicon load.

AMD Instinct MI300X Platform Data Sheet (GD-83, PID# 232405395-D) ↗
AMD · eight accelerator screening profilereported + derived

AMD Instinct MI325X Platform (Universal Baseboard, 8x OAM)

8GPUCPU2.0 TBaccelerator memory8.0 kWpower basis

Cooling · Not stated in the platform data sheet — no air/liquid designation is given for the 8x OAM UBB baseboard itself.

Memory / DIMMs · Not specified — this data sheet documents the 8x OAM GPU baseboard only; no host CPU/DIMM memory configuration is given.

Network · 8x PCIe Gen 5 x16 (128 GB/s) host I/O, one per GPU; intra-baseboard AMD Infinity Fabric mesh at 7x 128 GB/s per GPU, 896 GB/s peak aggregate bidirectional ring bandwidth.

This is a GPU-baseboard-only figure (8x OAM accelerators on a UBB 2.0), not a full-server or full-rack power draw. It excludes host CPU(s), system memory, storage, networking, fans, PSUs, and AC/DC conversion losses, so it is not suitable for facility-level PUE math on its own — only for screening accelerator-silicon load.

AI Data Sheet: AMD Instinct MI325X Platform (GD-83) ↗
AMD · other short labelmanufacturer reported

AMD Instinct MI300A APU (illustrative 4-APU node architecture)

1GPU1CPU128 GBaccelerator memorynot disclosedpower basis

Cooling · Air or liquid capable up to 550W Max TDP per socket; liquid-cooling-only above that, up to 760W Max TDP per socket.

Memory / DIMMs · 128 GB HBM3 unified memory per APU package, coherently shared between the integrated CPU cores and GPU compute units (not conventional host DIMMs). No separate host-memory/DIMM configuration is given since the data sheet covers only the APU package.

Network · Per APU: 8x 128 GB/s AMD Infinity Fabric interfaces (4 dedicated Infinity Fabric, 4 assignable to Infinity Fabric or PCIe Gen 5 x16); scale-out listed as 400 Gbps Ethernet or InfiniBand; two x4 NVMe storage interfaces and two USB interfaces per APU (device counts/models not stated).

No AMD-published platform/system total power figure exists for any multi-APU MI300A configuration in this primary source — only per-socket TDP is stated, and the document's own '4-APU' reference architecture carries no accompanying wattage. Third-party OEM systems (e.g., Supermicro 4-way MI300A servers, HPE Cray EX blades used in El Capitan) do publish their own node-level power figures, but those are vendor-integrator sources, not AMD primary sources, and are out of scope for this candidate. This entry is included to document that the search was made and no genuine AMD platform-power figure was found, per the 'unknowns stay unknown' policy.

AMD Instinct MI300A APU Data Sheet (GD-83) ↗
Google · other short labelreported + derived

Cloud TPU v4

1GPUCPU32 GBaccelerator memory0.2 kWpower basis

Cooling · Not stated on this page (no cooling method specified for this measured-power figure).

Memory / DIMMs · Not stated on this page. HBM2 is on-package (32 GiB, 1200 GBps); no separate DIMM/memory-channel detail given.

Network · Six inter-chip interconnect (ICI) links per chip enabling a 3D mesh/torus pod topology; x16 PCIe gen3 interface to host (direct connect), per the same document.

Chip-only figure, not a system/rack/pod total. Does not cover host servers, networking, or cooling. Not directly comparable to full-server or full-rack power figures used elsewhere in this dataset. Google has not published an equivalent figure for TPU v5e, v5p, v6e (Trillium), or TPU7x (Ironwood) in any official documentation or blog post found during this research.

TPU v4 | Google Cloud Documentation ↗
Google · other short labelreported + derived

Cloud TPU v3

1GPUCPU32 GBaccelerator memory0.2 kWpower basis

Cooling · Not stated on this page (no cooling method specified for this measured-power figure).

Memory / DIMMs · Not stated on this page. HBM2 is on-package (32 GiB, 900 GBps); no separate DIMM/memory-channel detail given.

Network · 2D torus pod interconnect topology per the same document; per-chip host-connect interface not detailed on this page.

Chip-only figure, not a system/rack/pod total. Does not cover host servers, networking, or cooling. Not directly comparable to full-server or full-rack power figures used elsewhere in this dataset. Google has not published an equivalent figure for TPU v5e, v5p, v6e (Trillium), or TPU7x (Ironwood) in any official documentation or blog post found during this research.

TPU v3 | Google Cloud Documentation ↗
Amazon Web Services (AWS) · pod slicereported + derived

Amazon EC2 Trn2 UltraServer

64GPUCPU6.0 TBaccelerator memorynot disclosedpower basis

Cooling · Not stated by AWS -- no cooling method or thermal design is published for the Trn2 UltraServer on the reviewed pages.

Memory / DIMMs · Not stated by AWS -- no host CPU/DIMM memory configuration is published for the Trn2 UltraServer on the reviewed pages.

Network · 64 Trainium2 chips connected via NeuronLink, delivering up to 83.2 FP8 petaflops of compute, 6 TB of total HBM3 with 185 TBps of total memory bandwidth, 12.8 Tbps of EFAv3 networking, and support for up to 32 TB of local NVMe storage.

Captures only the compute/memory/network specs AWS publishes for the Trn2 UltraServer. No power, thermal, cooling, DIMM, or physical host-CPU details are published at chip, instance, or aggregate UltraServer level. Local NVMe capacity (up to 32 TB) is stated but device count is not. Not usable for PUE or facility power planning without an independently sourced power figure.

Amazon EC2 Trn2 Instances — AWS (EC2 Trn2 product page) ↗
Amazon Web Services (AWS) · full serverreported + derived

Amazon EC2 Trn2 Instance (trn2.48xlarge)

16GPUCPU1.5 TBaccelerator memorynot disclosedpower basis

Cooling · Not stated by AWS -- no cooling method or thermal design is published for trn2.48xlarge on the reviewed pages.

Memory / DIMMs · AWS states 2 TB of total instance memory for trn2.48xlarge (see cpu_memory_gb) but does not publish DIMM count, type, or channel configuration.

Network · 16 Trainium2 chips deliver up to 20.8 FP8 petaflops of compute, 1.5 TB HBM3 total with 46 TBps memory bandwidth, and 3.2 Tbps of EFA networking; supports up to 8 TB local NVMe storage. The trn2.48xlarge instance has 192 vCPUs and 2 TB of instance memory.

No power, thermal, cooling, DIMM, or physical host-CPU (as opposed to vCPU) details are published by AWS for the Trn2 instance. Not usable for PUE or facility power planning without an independently sourced power figure.

Amazon EC2 Trn2 Instances — AWS (EC2 Trn2 product page) ↗
Amazon Web Services (AWS) · full serverreported + derived

Amazon EC2 Inf2 Instance (inf2.48xlarge)

12GPUCPU384 GBaccelerator memorynot disclosedpower basis

Cooling · Not stated by AWS -- no cooling method or thermal design is published for inf2.48xlarge on the reviewed pages.

Memory / DIMMs · AWS states 768 GiB of total instance memory for inf2.48xlarge (see cpu_memory_gb) but does not publish DIMM count, type, or channel configuration.

Network · inf2.48xlarge: 12 Inferentia2 chips, 384 GB total accelerator memory (32 GB per chip), 9.8 TB/s of total memory bandwidth, 192 GB/s NeuronLink interconnect between chips, 192 vCPUs, 768 GiB instance memory, and up to 100 Gbps network bandwidth.

No power, thermal, cooling, DIMM, or physical host-CPU (as opposed to vCPU) details are published by AWS for Inf2 instances. Not usable for PUE or facility power planning without an independently sourced power figure.

Amazon EC2 Inf2 Instances — AWS (EC2 Inf2 product page) ↗
Intel · oam acceleratormanufacturer reported

Intel Gaudi 3 AI Accelerator HL-325L OAM Mezzanine Card

1GPUCPU128 GBaccelerator memory0.9 kWpower basis

Cooling · Not specified in this product brief (no air vs. liquid cooling designation is given alongside the 900W figure).

Memory / DIMMs · Not applicable / not stated — this is an accelerator mezzanine card, no DIMM slots are documented on it.

Network · 9.6 Tbps bi-directional networking capacity via 24x200 GbE RoCE v2 RDMA ports integrated on-die (native on-chip RDMA over converged Ethernet).

Single-accelerator component TBP only, not a system/server/rack figure. Cooling method not stated. Not itself an 8-accelerator platform figure — see the companion HLB-325 baseboard candidate for that.

Product Brief: Intel Gaudi 3 AI Accelerator HL-325L OAM Mezzanine Card ↗
Intel · eight accelerator screening profilereported + derived

Intel Gaudi 3 AI Accelerator HLB-325 Baseboard (8x HL-325L OAM accelerators)

8GPUCPU1.0 TBaccelerator memory7.6 kWpower basis

Cooling · Not specified (no air vs. liquid cooling designation given). Brief states the baseboard 'derives the majority of its power from 54V power input, only requiring a separate 12V input for standby.'

Memory / DIMMs · Not applicable / not stated — the baseboard carries 8 accelerator OAM cards, not CPU DIMMs.

Network · Baseboard provides 4.2 TB/s bi-directional all-to-all bandwidth directly between the 8 onboard accelerators (no separate switching IC required), plus an additional 1.2 TB/s bi-directional scale-out bandwidth via 6 OSFP connectors (24x200 Gbps RoCE v2 RDMA).

This is a baseboard/accelerator-tray-only power figure explicitly excluding the host CPU subsystem, system memory, storage, and chassis cooling — it is not a complete server or rack figure and should not be used directly for facility-level PUE math without adding host-server overhead. No cooling method (air/liquid) is stated.

Product Brief: Intel Gaudi 3 AI Accelerator HLB-325 Baseboard ↗
Intel · full servermanufacturer reported

Intel HLS-Gaudi2 AI Accelerator Server

8GPU2CPU768 GBaccelerator memory7.5 kWpower basis

Cooling · Not explicitly stated in the datasheet (no air vs. liquid cooling designation given); described only as a 19-inch rack-mount server chassis.

Memory / DIMMs · 1TB of host system memory stated for the dual-socket Xeon 8380 subsystem ('Dual socket Intel 3rd Generation Xeon 8380 CPUs with 1TB memory'); DIMM count/type not specified. Separately, 768GB of HBM2E is on-package accelerator memory across the 8 Gaudi 2 cards (96GB each, stated directly in the spec table).

Network · 24x100GbE RoCE v2 RDMA scale-out via 6x QSFP-DD, plus internal non-blocking all-to-all interconnect (21x100GbE RoCE ports from each accelerator to the other 7), plus two 2x100GbE PCIe host NICs for host connectivity (per block diagram description).

Power figure is Intel's stated 'Max Power Usage,' likely a design/nameplate ceiling rather than a continuously-measured typical draw; cooling method, PSU redundancy, and AC/DC conversion overhead are not stated. This document describes Gaudi 2 (not Gaudi 3) and is dated October 2023 per its filename/copyright; hosted on habana.ai, Intel's wholly-owned AI-accelerator subsidiary domain, and is Intel-branded throughout ('Intel® HLS-Gaudi®2'). An Intel-hosted landing page for this same datasheet exists at intel.com/content-details/784779 but returned HTTP 403 when fetched directly.

Intel HLS-Gaudi2 AI Accelerator Server datasheet ('AI Performance with Ethernet Scale') ↗
Cerebras · full servermanufacturer reported

CS-3 (Cerebras Wafer-Scale System, 1x WSE-3)

1GPUCPU44 GBaccelerator memory27.0 kWpower basis

Cooling · Liquid cooling, direct-to-chip. Datasheet specifies: Water Filtering 50 micron; Water Chemistry PG25; Water Flow Rate 100 L/min/system; Water Temperature 20 ± 2°C.

Memory / DIMMs · Not stated. The datasheet lists 44 GB / 21 PB/s on-chip SRAM as processor memory; no host DIMM/DRAM configuration is given for the CS-3 system itself.

Network · 1.2 Tb/s I/O bandwidth; management: 1x 1GbE management port (RJ45), 1x console port (RJ45), 1x PoE flow controller interface.

Single-system power draw only; excludes any external MemoryX/SwarmX or networking cabinets used in multi-system training/inference clusters (the datasheet's separate 'Inference Cluster Configurations' table gives multi-cabinet provisioned heat loads instead, from ~240 kW for an 8-node/5-cabinet cluster up to multi-MW for 64-88 node clusters). No PUE/facility overhead included. CPU/host memory/storage not specified by this datasheet.

Cerebras 'CS-3 System Specifications' datasheet (PDF), linked directly from cerebras.ai/system ↗
Cerebras · full rackreported + derived

CS-3 Rack (2x Cerebras CS-3 systems)

2GPUCPU88 GBaccelerator memory54.0 kWpower basis

Cooling · Combined liquid + air cooling. Datasheet states 'Water Cooling Load per Rack: 50.4 kW' and 'Air Cooling Load per Rack: 3.6 kW' (sums to the 54 kW total). Liquid cooling water coupling: 1.5" sanitary fittings, 2 pairs supply/return. Power feed: 2+1 redundant, 240V/415Y 3-phase, 60A, 5-wire, IEC 60309 plug.

Memory / DIMMs · Not stated.

Network · 8x QSFP-DD transceiver, 4x100G; management network: 1x 1GbE management port (RJ45), 1x console port (RJ45), 1x PoE flow controller interface.

Rack-level total for exactly 2x CS-3 systems only; excludes any additional WSE NET/interconnect cabinets used in larger multi-rack clusters (the datasheet's separate cluster-configuration table shows those as additional cabinets with their own heat loads, e.g. ~240 kW provisioned total for an 8-node/5-cabinet cluster). No PUE/facility overhead included.

Cerebras 'CS-3 Rack Specifications' datasheet (PDF), linked directly from cerebras.ai/system ↗
Groq · pcie acceleratorreported + derived

GroqCard Accelerator (GC1-010B)

1GPUCPU0 GBaccelerator memory0.4 kWpower basis

Cooling · Not stated in this spec sheet. GroqCard is a PCIe Gen4 x16 add-in card with no dedicated cooling spec published here; it relies on host-server chassis airflow.

Memory / DIMMs · Not applicable / not stated. GroqCard has no DIMM slots — its 230 MB of memory is on-die SRAM, not DRAM.

Network · Up to 11 RealScale chip-to-chip connectors for multi-server/multi-rack scalability without external switches; PCIe Gen4 x16 host interface, up to 31.5 GB/s bi-directional bandwidth.

Component-level (single PCIe accelerator card) power figure only — not a system, node, or rack total. Groq's GroqNode server datasheet states it packages up to 8x this card plus dual host CPUs (2x AMD EPYC 7313) and 1TB DRAM in a 4U chassis with '4 x 2000W (220-240VAC)' power supplies, but that is a PSU nameplate-capacity figure (likely N+redundant), not a stated power-draw number, so no full-node or full-rack GroqNode/GroqRack power figure is included here. We also could not locate a working GroqRack spec PDF with any published rack-level power figure as of this check.

GroqCard™ Accelerator Product Spec Sheet, v1.5 (PDF), linked from groq.com/groqcard-accelerator ↗
Huawei · full serverreported + derived

Atlas 800T A2 Training Server

GPU4CPUaccelerator memory5.8 kWpower basis

Cooling · Air-cooled; 8 hot-swappable fan modules in N+1 redundancy mode (per Hardware Description, Maintenance and Service Guide 05).

Memory / DIMMs · Up to 32 DDR4 DIMM slots (RDIMM), max speed 3200 Mbit/s, ECC/SEC-DED/SDDC + patrol scrubbing protected, 16 GB/32 GB/64 GB per module (all installed modules must share the same part number).

Network · CPU mainboard supports one FlexIO card providing 4x 25GE/10GE optical ports (PXE-capable); NPU carrier board supports up to 4 parameter-plane interface cards, each with 2x 200GE optical ports (fiber or copper).

Does not capture NPU (Ascend 910) accelerator count or exact model suffix (e.g. 910B) -- Huawei's own Overview page names the processor only generically as 'Ascend 910 AI Processors' without a quantity on the pages retrieved, so gpu_qty is left null and gpu_model reflects only the generic name actually published. Power figure is a manufacturer-stated maximum across configurations, not a measured or typical operating draw. Storage device count and accelerator HBM capacity are not stated on the retrieved pages.

Atlas 800T A2 Training Server User Guide 04 (Huawei Enterprise Support) — Overview, Technical Specifications, and Physical Specifications sections ↗
Huawei · full servermanufacturer reported

Atlas 800I A3

8GPU4CPUaccelerator memory15.4 kWpower basis

Cooling · Air-cooled; deployable in standard air-cooled equipment rooms, using integrated fan modules plus Huawei-developed liquid-assisted air cooling (LAAC) modules. Each drawer integrates 5 fan modules with 4+1 hot-swap redundancy (single-fan failover time <= 2 minutes).

Memory / DIMMs · Not stated with specific capacity or DIMM type on the pages retrieved.

Network · Not stated with explicit port-count/speed figures on the pages retrieved. The server integrates separate 'LingQu UB' interconnect boards, documented by Huawei in a distinct 'LingQu UB Device Hardware' / 'LingQu UB Device Specifications' section, for supernode-to-supernode fabric connectivity.

Does not capture accelerator HBM memory capacity or interconnect bandwidth -- figures such as '128 GB HBM per NPU' or specific 400GE/UB link counts are widely reported by industry/trade press but were not found stated in these terms on the specific Huawei pages retrieved, so accelerator_memory_gb and detailed network specs are left null. The exact Ascend NPU variant (e.g. '910C') is not named as such in the retrieved Huawei text, which refers to it generically as an 'Ascend 910 AI module' (HiAM module). Power figure is a manufacturer-stated maximum across configurations, not a measured or typical operating draw, and excludes the separate LingQu UB Device module's own power draw.

Atlas 800I A3 User Guide 04 (Huawei Enterprise Support) — Product Features and Physical Specifications sections ↗

Infrastructure intelligence atlas primary · secondary · tertiary · buyer · producer

A denser layer than MW alone: power-train stages, cooling topology bands, rack-density eras, official buyer/producer indexes, and equipment lead-time stress tests. Screening orientation only — manufacturer BOMs and site engineering still win procurement.

5power stages 4cooling bands 8market indexes 5lead-time screens 39source feeds

Primary → tertiary power train

Primary power

Utility delivery

Transmission/distribution interconnection, customer substation, utility transformers, metering, protection, curtailment terms

Units · MVA · kV · fault MVA · energization date

RFP hooks · service voltage · firm vs interruptible · deposit & study fees · LLI milestones · N-1 utility contingency

ISO/RTO queues · utility large-load tariffs · EIA-930 system context

Primary power

Campus MV distribution

MV switchgear, feeders, bus duct, main-tie-main, on-site transformers to LV

Units · 15–34.5 kV gear · MVA transformers · %Z

RFP hooks · lead-time holds · arc-flash · maintainability · approved equals · spare transformers

DOE distribution-transformer rules · ENERGY STAR UPS adjacency · AHRI cooling power draw

Secondary power

UPS + static transfer

UPS topology, battery/flywheel autonomy, STS, PDU/RPP, busway to rack

Units · kW/kVA · efficiency@%load · minutes autonomy

RFP hooks · part-load efficiency curve · redundancy boundary · battery chemistry · bypass maintainability

ENERGY STAR UPS directory

Secondary power

Rack / busbar power

Busway, busbar, power shelves, PSU conversion, HVDC experiments (OCP)

Units · kW/rack · Vdc/Vac · conversion stages

RFP hooks · rack density roadmap · hot-aisle/liquid fraction · PSU redundancy · harmonics

NVIDIA/AMD primary BOMs · OCP rack specs

Tertiary power

Backup generation + BESS

Diesel/HVO/gas gensets, fuel storage, paralleling, BESS, black start, emissions controls

Units · MW blocks · hours autonomy · g/bhp-hr emissions class

RFP hooks · air permit family · noise ordinance · fuel logistics · load-step tests

EPA engine certification data · state air permits

Cooling topology vs rack density

Cooling topology

Traditional air (CRAH/CRAC + raised floor / containment)

Rack band · ≤ ~15–20 kW/rack typical; practical ceiling often cited ~30–40 kW with rear-door assist

PUE band · ~1.3–1.6 (design- and climate-dependent)

Water · Often cooling-tower or dry-cooler dependent; WUE is site-specific

Use when · Legacy enterprise, moderate-density colo, inference pods that stay air-coolable

Industry practice band for screening; validate with ASHRAE thermal guidelines and OEM envelopes

Cooling topology

Rear-door / in-row heat exchangers

Rack band · ~20–40 kW/rack common screening band

PUE band · ~1.2–1.4 when free cooling is available

Water · Facility water loop; can reduce air-handler count

Use when · Mid-density GPU refresh inside existing shells

Industry practice band; pair with AHRI datacom ratings for equipment shortlists

Cooling topology

Direct-to-chip liquid cooling (DLC / cold plate)

Rack band · ~40–150 kW/rack; AI training racks often designed 100–140+ kW

PUE band · ~1.03–1.15 when liquid fraction is high

Water · Secondary loop flow is not water consumption; WUE still needs a water balance

Use when · GB200/GB300-class racks and dense OAM platforms

Aligned with public AI-rack power disclosures (e.g. NVL72 up to 142 kW) plus industry DLC practice

Cooling topology

Immersion (single-phase or two-phase)

Rack band · ~100–200+ kW/rack screening band

PUE band · ~1.02–1.08 claimed in vendor case literature — treat as aspirational until measured

Water · Dielectric fluid inventory, leak detection, and heat-rejection water still apply

Use when · Extreme density experiments, specialized HPC, or shells designed around tanks

Secondary/vendor literature heavy — require FAT/SAT and third-party thermal tests in RFPs

Rack density timeline

EraRack powerNoteEvidence class
Pre-AI enterprise (illustrative)4–12 kWHistorical average enterprise densities often cited near mid-single to low-double-digit kW/rack.secondary_industry
Modern air-cooled GPU pods20–40 kWMany retrofit halls target this band before full liquid conversion.secondary_industry
NVIDIA GB200/GB300 NVL72-class AI rackup to 142 kW (GB300 NVL72 manufacturer maximum)NVIDIA's GB300 NVL72 Enterprise Reference Architecture states up to 142 kW for the full liquid-cooled rack — a manufacturer maximum, not a workload average.primary_manufacturer
Next-wave liquid AI cabinets (roadmap / secondary)200–600+ kW discussed in industry and vendor roadmapsTreat roadmap kW figures as scenario inputs, not purchaseable nameplate, until an OEM publishes a BOM.secondary_roadmap

Buyer & producer indexes

buyer index

U.S. private data-center construction (Census VIP)

Metric · Seasonally adjusted annual-rate $ put in place

May 2026 preliminary · ~$59.3B SAAR (secondary rebroadcast of Census private construction detail)

Monthly

Cleanest official monthly dollar flow into U.S. data-center construction — a buyer CapEx index.

Open series ↗
buyer index

Public tender / RFP velocity (SAM.gov + EU TED)

Metric · Open notices for servers, UPS, generators, cooling, construction, ops

Continuous — filter by NAICS/PSC/CPV + place of performance

Daily / near-daily

Buying intent before awards — the language of scopes, options, and delivery dates.

Open series ↗
producer index

Transformer producer price (BLS PCU335311335311)

Metric · Index level, MoM and YoY %

Track via BLS/FRED; May 2026 prints near ~457 (index, not $)

Monthly

Producer price pressure on the longest-lead electrical equipment.

Open series ↗
producer index

Server / host-computer producer price (WPU11510116)

Metric · Index level for multiuser host computers

Monthly BLS/FRED series

Monthly

Hardware inflation/deflation input for server BOMs and lease pricing models.

Open series ↗
buyer index

Data processing & hosting service prices (NAICS 518210)

Metric · Industry PPI for hosting / IT infrastructure provisioning

Monthly BLS industry product series

Monthly

Service-side price index for colo/cloud-like outputs.

Open series ↗
producer index

Taiwan ODM / semiconductor monthly revenue (TWSE)

Metric · Listed-company monthly revenue

High-frequency producer throughput for AI servers and silicon

Monthly (company filings)

Producer supply heartbeat before annual reports — boards, power, cooling, and chips.

Open series ↗
producer index

U.S. import/export flows (USITC DataWeb)

Metric · Value and quantity by HTS for transformers, generators, chillers, servers, optics

Monthly merchandise trade

Monthly

Physical supply flows and tariff exposure for the equipment bill of materials.

Open series ↗
buyer index

Global data-centre TWh envelope (IEA)

Metric · TWh/year scenarios

Base case ~945 TWh by 2030 (~3% of world electricity)

Report cycles

Global demand ceiling used by policymakers and utilities; not a project queue.

Open series ↗

Lead-time stress tests

Secondary survey and industry bands for schedule risk — not purchase-order commitments. Confirm with current OEM quotes.

EquipmentBandEvidenceNotePlanning use
Large power transformers~100–144+ weeks commonly reported; some GSU quotes longersecondary_surveyWood Mackenzie Q2 2025 survey figures widely reported: ~128 weeks power transformers, ~144 weeks GSUs. Use as a schedule stress test, not a purchase order.Order at concept / before final design freeze; dual-source and accept-alternates language in RFP
MV switchgear~44–80 weeks reported depending on class and customizationsecondary_surveyTrade-press summaries of 2025–2026 utility surveys; confirm with current OEM quotes.Freeze one-line early; hold production slots
Diesel / large standby generators~50–78 weeks for many 1.5–3 MW-class blocks (indicative)secondary_industryHighly configuration-dependent; emissions aftertreatment and paralleling gear add time.Air-permit family and fuel strategy lock with procurement
AI accelerator supply (GPU/OAM)Allocation-driven; not a single public lead-time seriessecondary_marketTrack producer revenue (TWSE), foundry revenue, and OEM allocation letters rather than a BLS lead-time series.BOM revision control + dual architecture options in RFP
Liquid-cooling CDUs / facility water skidsProject-specific; often on critical path once rack density exceeds air limitssecondary_industryPair AHRI datacom certifications with factory-witness tests; no single public national lead-time index.Specify ΔT, approach, water quality, and controls interoperability explicitly
ERCOT deep dive

Watch feed

2026-07-24 batch zero deadline

Batch Zero Load Information Form due to ERCOT

Interconnecting TSPs must submit the Batch Zero LIF workbook, Dynamic Stability Study Form, and Section 9.2.1.2/9.2.1.4 attestations to BatchZero@ercot.com source

2026-07-10 batch zero deadline

Form X / Form W / most ILLE attestations due

ILLEs must submit Form X (WLPUN designation), Form W (PCLR commitment), and Section 9.2.1.1 attestations to their interconnecting DSP/TSP source

2026-06-18 regulatory

PUCT approves Batch Zero rules

PUCT approved rules for ERCOT to review an initial batch of large-load interconnection requests (75+ MW), including a $50,000/MW fee structure source

2026-03-31 queue update

Large-load queue reaches 438 GW

ERCOT tracking 438 GW of large-load requests; ~90% from data centers; 198 GW applied in Q1 2026 alone source

2026-03-13 queue update

137 new LLI submissions totaling ~140 GW by 2036

ERCOT reported receiving 137 new LLI submissions totaling approximately 140,000 MW of new large load by 2036, still being processed source

2025-11-30 queue update

Queue nearly quadruples year-over-year

226 GW of large-load interconnection requests, up from 63 GW twelve months earlier; 73% from data centers source

Snapshot table

MonthTotal queueApprovedObserved peakSource
2026-03-31438.0 GW9,042 MW3,883 MWUtility Dive / PUCT; ERCOT Mar 2026 TAC report
2025-11-30226.0 GWERCOT public data via davefriedman.substack.com
2025-03-319,042 MW3,883 MWERCOT LLI Status Update Mar 2026 (TAC)
2024-11-3063.0 GWERCOT public data via davefriedman.substack.com
2024-01-3139.3 GW4,479 MW2,587 MWERCOT LLI Status Update Jan 2024
2023-12-3139.2 GW4,479 MW2,432 MWERCOT LLI Status Update Jan 2024
2023-10-3139.3 GW3,926 MW2,379 MWERCOT LLI Status Update Jan 2024
2023-09-3039.6 GW3,926 MW2,379 MWERCOT LLI Status Update Jan 2024
2023-08-3142.5 GW3,744 MW2,347 MWERCOT LLI Status Update Jan 2024
2023-07-313,084 MW2,336 MWERCOT LLI Status Update Jan 2024
2023-05-3141.8 GW2,620 MW2,072 MWERCOT LLI Status Update Jan 2024
2023-04-3039.6 GW2,570 MW2,013 MWERCOT LLI Status Update Jan 2024
2023-03-3139.4 GW2,570 MWERCOT LLI Status Update Jan 2024
2022-04-3017.4 GWERCOT LLI Status Update Jan 2024

Compute per megawatt worldwide · TOP500 flagship systems

Dual axis · log efficiency (Gflops/W) + HPL power (MW) · measured as published, never estimated

Flagship TOP500 efficiency improved 134× from 2008 to 2024 — a doubling every ~2.3 years. The ERCOT large-load queue nearly doubled in the last twelve months alone. Efficiency isn't going to save the grid: today's 438.0 GW queue equals ~14,807 copies of El Capitan, the most powerful computer on the TOP500.

ListSystemCountryHPL RmaxPowerGflops/W
2024-11El CapitanUnited States1,742.0 PF29.6 MW58.89
2024-06AuroraUnited States1,012.0 PF38.7 MW26.15
2022-06FrontierUnited States1,102.0 PF21.1 MW52.23
2020-11FugakuJapan442.0 PF29.9 MW14.78
2019-06SummitUnited States148.6 PF10.1 MW14.72
2016-06Sunway TaihuLightChina93.0 PF15.4 MW6.05
2013-06Tianhe-2China33.9 PF17.8 MW1.90
2012-11TitanUnited States17.6 PF8.2 MW2.14
2011-11K computerJapan10.5 PF12.7 MW0.83
2009-11JaguarUnited States1.8 PF7.0 MW0.25
2008-06RoadrunnerUnited States1.0 PF2.3 MW0.44

Extreme-environment compute ledger orbit · ocean · floating · underground · atmosphere · polar

Documented data centers, computing payloads, and proposals in non-standard environments — including Antarctic research IT (McMurdo IT&C, IceCube, South Pole Telescope). Communications platforms and edge systems are labelled when they are not full data centers; no unpublished MW is inferred. Commercial hyperscale does not operate on the Antarctic continent under the Treaty System.

deployed commercial cloud deviceorbit · LEO

AWS Snowcone on Ax-1

Axiom Space · International Space Station

Commercial cloud device operationalized aboard the ISS in 2022; commercial AI inference conducted in orbit. Edge device, not a full data center.

As of 2022-04 · Axiom Space

deployed research prototypeorbit · LEO

AxDCU-1

Axiom Space · International Space Station

Data Center Unit-1 powered by Red Hat Device Edge; cloud computing, AI/ML, data fusion, and space-cybersecurity test applications.

As of 2025 fall · Axiom Space

deployed commercial nodesorbit · LEO

Orbital Data Center nodes 1–2

Axiom Space · Free-flying low Earth orbit nodes

First two dedicated orbital data-center nodes launched January 11, 2026; 2.5 Gbps-capable optical intersatellite links.

As of 2026-01 · Axiom Space

deployed research systemorbit · LEO

Spaceborne Computer-2

NASA + Hewlett Packard Enterprise · International Space Station

HPE ProLiant DL360 server plus Edgeline 4000 server; research and commercial edge-compute validation.

As of 2025-02 · NASA Spinoff

deployed technology demonstratororbit · LEO

Starcloud-1

Starcloud · Low Earth orbit

First NVIDIA H100 GPU in orbit; company reports Gemini inference and nanoGPT training in space.

As of 2025-11 · Starcloud

planned orbital nodeorbit · LEO

AxODC Node ISS

Axiom Space · International Space Station

Planned additional ISS ODC node; Axiom says it plans at least three interconnected ODC nodes by 2027.

As of 2027 target · Axiom Space

planned commercial missionorbit · LEO / SSO

Starcloud-2

Starcloud · Sun-synchronous orbit

GPU cluster; persistent storage; 24/7 access; proprietary thermal and power systems in a smallsat form factor.

As of 2027 target · Starcloud

studyorbit · feasibility study

Space-Based Data Centres study

European Space Agency · ESA Discovery and Preparation study

Generic space-data-centre architecture and simulation; the work examines technical operational and economic constraints before mission approval.

As of 2024-05 · ESA Discovery and Preparation

Worldwide connection & buildout sources 105 primary and authoritative feeds · buyer + producer + physical system

Project queues, product certification, server/UPS/storage/network BOMs, water and emissions, tenders, trade, construction, supplier revenue, climate, and physical risk. Each card states its signal, cadence, access path, and authority so a forecast is never mistaken for a transaction or a certified model.

39new research feeds1backup power2buyer index1climate4compute1construction demand1cooling3environmental permits1equipment compliance3facility performance1grid carbon4grid demand2grid pipeline1network1physical risk2power2procurement3producer index1project pipeline1storage1supplier demand1trade2water
air permitsVirginia

Virginia DEQ issued air permits for data centers

Statewide registry of every air permit issued to a data center in Virginia (backup diesel-generator permits): site name, registration number, issuance date, program type and county, with 198 rows as of July 6, 2026. Bot-blocked; scraping requires a browser-like session.

Virginia Department of Environmental Quality · facility-level, per permit

updated roughly monthly with explicit as-of date · public

atmospheric compute studyAtmosphere / stratosphere

Data Center-Enabled High Altitude Platforms research

Research proposal for a flying data-center-enabled high-altitude platform, including energy, solar, thermal, communication, and offloading analysis. It is a study, not a deployment record.

Academic research proposal · Academic system proposal and model

Published research; no commercial deployment identified · public

atmospheric edgeStratosphere / global

Airbus / AALTO Zephyr HAPS

Primary platform source for the solar-electric Zephyr HAPS operating above 60,000 feet with earth-observation and connectivity payloads. It is tracked as an airborne compute-adjacent platform, not a data center.

Airbus / AALTO · Aircraft / stratospheric payload platform

Months-long endurance demonstrated; commercial service and payload status evolve · public

atmospheric edgeStratosphere / Japan

SoftBank HAPS / Sceye stratospheric platform

Primary HAPS source covering Sunglider testing at 19 km, Sceye long-duration platform, solar power, payloads, and planned pre-commercial communications service. HAPS is tracked as atmospheric edge infrastructure, not a full data center.

SoftBank Corp. · Aircraft / HAPS platform and communications payload

Flight-tested; pre-commercial service targeted for Japan in 2026 · public

backup_powerVehicles, engines, and equipment subject to U.S. EPA certification

EPA Annual Engine and Equipment Certification Data

Lets backup-generator RFPs connect engine families to certified emissions characteristics before air-permit screening.

U.S. Environmental Protection Agency · Certified engine families, manufacturers, fuels, power classes, emissions standards, and compliance attributes.

Interactive data updated daily; downloadable files generally quarterly · Public dashboard and spreadsheet downloads

buyer_indexU.S. data processing, hosting, and related services

BLS Producer Price Index — Data Processing, Hosting & Related Services (NAICS 518210)

Closest official price index for the service side of the data-center product — a buyer/seller contract benchmark, not a facility MW measure.

U.S. Bureau of Labor Statistics · Producer prices received by data-processing and hosting establishments, including IT infrastructure provisioning.

Monthly · BLS PPI industry tables and FRED industry product series (e.g. PCU5182105182105)

buyer_indexUnited States macro and industry series

FRED Economic Data — construction, PPI, and server series mirrors

Practical API layer for stitching construction-spend (buyer), transformer PPI (producer), and server PPI into one monitoring graph.

Federal Reserve Bank of St. Louis (mirrors official agency series) · Machine-readable mirrors of Census VIP, BLS PPI, and related industrial series used in buyer/producer dashboards.

Follows source agency release calendars · Public FRED API and series pages

capacity + congestionNetherlands

Netbeheer Nederland national capacity map + large-user queue

National capacity map for large users: available transport capacity, congestion, queue counts, and requested MW by supply area. It is a constraint and connection-readiness source, not a data-center project list.

Netbeheer Nederland · Feeding area / municipality; large-consumer connection capacity

Map and queue data updated monthly · public

capacity pricesPJM (13 US states + DC)

PJM RPM Base Residual Auction results

Official PJM capacity-market auction library with results, supply curves and planning parameters for every delivery year back to 2007/08. The 2027/2028 BRA cleared at $333.44/MW-day RTO-wide (up from $329.17), with total cleared-capacity cost of $16.4B.

PJM Interconnection · per-auction, by locational deliverability area and resource type

annual auctions with per-auction reports · public

capex disclosureUS-listed companies (global operations)

SEC EDGAR XBRL API — hyperscaler and DC-REIT capex

Free JSON APIs over every SEC filing: the XBRL company-concept endpoint returns full quarterly time series of tagged line items such as purchases of property and equipment for Microsoft, Alphabet, Amazon, Meta and the data-center REITs. Filings appear within minutes of release; a descriptive User-Agent header is required.

U.S. Securities and Exchange Commission · company, quarterly line item

real-time as filed (quarterly for earnings) · public API

capex trackerglobal

Synergy Research hyperscale capex and capacity articles

Paid-tracker firm publishing headline numbers free in its article stream: hyperscale operator capex of $142B in Q3 2025, 1,297 hyperscale data centers worldwide with 770 in pipeline, and US share of hyperscale capacity at 55%.

Synergy Research Group · global aggregates by operator class; quarterly data points

several articles per month, quarterly after earnings season · public summary

climateUnited States and global observing stations, dataset-dependent

NOAA Climate Data Online Web Services

Supports free-cooling hours, design-day temperature, evaporative-water stress, flood/precipitation, and long-run climate-normal comparisons.

NOAA National Centers for Environmental Information · Station-level temperature, precipitation, normals, extremes, and weather observations.

Daily observations with periodic quality control and normals updates · Public token-authenticated REST API and downloads

computeAMD Instinct MI355X accelerator

AMD Instinct MI355X Specifications

Enables transparent GPU-only load and memory arithmetic while exposing what a full platform RFP still must specify.

AMD · OAM accelerator TBP, HBM3E capacity/bandwidth, host interface, and thermal implementation options.

Manufacturer product lifecycle · Public product specification page

computeENERGY STAR-certified enterprise server models sold in the United States

ENERGY STAR Certified Enterprise Servers

Turns CPU, memory, PSU, and measured efficiency claims into a comparable machine-readable buying universe.

U.S. Environmental Protection Agency · Model-level processor/socket, ECC memory, power-supply, idle-power, active-efficiency, and configuration records.

Continuously maintained product registry · Socrata API and CSV export

computeGB300 NVL72 AI factory reference design

NVIDIA GB300 NVL72 Enterprise Reference Architecture

Provides a defensible bridge from AI capacity targets to rack count, memory, network ports, liquid cooling, and electrical scope.

NVIDIA · Rack and tray BOM: GPU/CPU counts, HBM and CPU memory, NVMe, network adapters, power shelves, cooling, and rack maximum power.

Versioned architecture documentation · Public HTML documentation

computeOpen rack standards adopted by large cloud operators

Open Compute Project — Orv3 / high-power rack designs

Bridges component BOMs to facility busway, busbar, and HVDC design language used in hyperscale RFPs.

Open Compute Project community (Meta and other contributors) · Open rack mechanical/electrical specs for high-power AI cabinets, including community HVDC directions.

Specification and summit release cycles · Public OCP contribution library and summit materials

connection capacityGreat Britain

National Grid ESO / NESO Connections 360

Maps signed connection projects and high-/low-voltage sites with contracted capacity. It can inform location and queue diligence for large loads, but it does not publish a data-center-specific queue and access requires registration.

National Energy System Operator (NESO) · Site / contracted-project connection data; Great Britain

Continuously updated as contracted projects are processed · registration required

construction spendingUnited States

Census Construction Put in Place — data center line item

Official monthly survey of US construction spending with a dedicated private 'Data center' line, published as machine-readable XLSX (seasonally adjusted annual rate plus unadjusted history). May 2026 preliminary shows $59.3B SAAR, exceeding all general office construction.

U.S. Census Bureau · national line item, monthly USD millions (SAAR and NSA)

monthly (~1st business day, two-month lag) · public

construction_demandUnited States

U.S. Census Construction Spending (VIP) API

Provides the clearest official monthly buyer-side index of dollars flowing into U.S. data-centre construction.

U.S. Census Bureau · Monthly value of construction put in place, including the explicit private data-centre construction series.

Monthly · Public Census timeseries API; API key policy applies

coolingParticipating datacom cooling product families

AHRI Datacom Cooling Certification Directory

Moves cooling RFPs from marketing claims toward comparable, third-party-certified thermal performance.

Air-Conditioning, Heating, and Refrigeration Institute · Independently certified capacities and ratings for datacom cooling equipment under the DCOM program.

Continuously maintained certification program · Public certification program and directory links

cooling surveyglobal

Uptime Institute Cooling Systems Survey

Annual survey (1,033 respondents in 2025) of data-center cooling technology mix, including direct liquid cooling adoption at 22% of organizations, DLC drivers and barriers, and the rack-density threshold at which operators expect air cooling to fail.

Uptime Institute · industry-wide survey shares, some regional/sector crosstabs

annual · public summary

crypto load proxyworldwide, with US state breakdown

Cambridge Bitcoin Electricity Consumption Index (CBECI)

Daily model-based estimate of Bitcoin network electricity consumption with a free CSV download API (134.19 TWh annualised best guess as of 2026-07-11) and a companion mining map with country and US-state hashrate shares. CC BY-NC-SA licensed.

Cambridge Centre for Alternative Finance, University of Cambridge · daily min/max/best-guess annualised TWh; mining map by country and US state

daily · public API

data-center connectionIreland

EirGrid data-center demand connections (DCCOPP v3)

Official data-center connection policy, application form, technical assessment, constrained-area criteria, demand fees, maximum-import-capacity process, and energisation prerequisites. Individual applications are not published as a public queue.

EirGrid / SONI · Data-center demand connection process; transmission and qualifying distribution demand

Policy and application material revised as CRU directions change · public

demand projectionglobal (US/China/Europe breakouts)

IEA Energy and AI report

IEA's dedicated study of data-centre and AI electricity demand: ~415 TWh consumed globally in 2024 (~1.5% of world electricity), projected to ~945 TWh by 2030, with US at 45% of 2024 consumption and China at 25%. Underlying dataset downloadable from the report page.

International Energy Agency · global and regional TWh, scenario-based to 2030/2035

flagship report with annual refreshes via World Energy Outlook · public

demand responseTexas (ERCOT)

ERCOT Emergency Response Service procurement results

ERCOT's paid emergency curtailment program in which qualified loads commit MW deliverable in 10 or 30 minutes, procured four times per year; awarded MW and cost by service type are posted per contract term via the NP3-144-M data product.

ERCOT · per contract term by service type and time period, awarded MW and cost

4x per year · public

efficiency registryEuropean Union

EU Energy Performance of Data Centres reporting scheme (EED Art. 12)

Mandatory EU-wide reporting scheme requiring every data centre with >=500 kW IT demand to report PUE, WUE, energy-reuse and renewable factors to a European database each May, with an aggregated public dashboard and periodic assessment report. Commission reporting puts EU data-centre electricity use at 70 TWh in 2024.

European Commission, DG Energy · per-facility reporting, published as EU/member-state aggregates

annual (reports due 15 May; dashboard updates follow) · public summary

energy modelUnited States

LBNL 2024 United States Data Center Energy Usage Report

DOE-commissioned bottom-up model of US data-center electricity use: 176 TWh in 2023 (4.4% of US electricity), projected to 325-580 TWh (6.7-12%) by 2028, with server/storage/network/cooling component breakdowns. Companion free benchmarking tools (DC Pro, PUE estimators) hosted on the same site.

Lawrence Berkeley National Laboratory / U.S. DOE · national totals with component breakdown and scenario ranges

periodic report (latest Dec 2024) · public

environmental_permitsU.S. facilities above reporting thresholds

EPA GHGRP / Facility Level Information on Greenhouse Gases Tool

Adds a carbon and industrial-neighbor layer for siting diligence beyond eGRID subregion averages.

U.S. Environmental Protection Agency · Large-facility greenhouse-gas reporting for power plants and industrial emitters near candidate campuses.

Annual reporting with periodic tool updates · Public data downloads and facility tools

environmental_permitsU.S. federally regulated facilities and delegated state-program records

EPA ECHO Web Services

Creates a repeatable environmental diligence layer around candidate campuses, generators, cooling discharge, and nearby industrial burden.

U.S. Environmental Protection Agency · Facility-level air, water, waste, inspection, enforcement, and compliance records.

Agency-system refreshes; varies by program · Public JSON/XML web services and bulk downloads

environmental_permitsVirginia data-center air permits

Virginia DEQ Issued Data-Center Air Permits

Leading development signal that appears before many commercial inventory databases — one permit is not one campus.

Virginia Department of Environmental Quality · Issued air permits for data-center facilities (backup generation and related emissions units).

Administrative updates as permits issue · Public permit list / agency page

equipment trackerglobal

Dell'Oro Group press room — data center infrastructure trackers

Free press releases carrying hard numbers from paid quarterly trackers: data center physical infrastructure (power/cooling) revenue ~$12B in 1Q 2026 (+28% YoY, thermal management +~50%), and data center semiconductor/component revenue +116% YoY.

Dell'Oro Group · market segment (power, cooling, switches, semis) per quarter

multiple releases per month, quarterly per tracker · public summary

equipment_complianceFederally regulated commercial and industrial equipment

DOE Compliance Certification Database

A procurement-grade cross-check for whether specified power and mechanical equipment has filed required federal certifications.

U.S. Department of Energy · Manufacturer-submitted certification reports for regulated equipment, including efficiency and model-family declarations.

Rolling manufacturer submissions · Public certification database and downloadable reports

facility pipelineVirginia

Piedmont Environmental Council Virginia data centers map

Continuously maintained ArcGIS web map of existing, under-construction and proposed data centers across Virginia, compiled from county land-use filings, with a companion diesel-generator air-quality map. A regional nonprofit tracker, but widely cited as the de-facto state dataset.

Piedmont Environmental Council · parcel/project-level (existing / under construction / proposed)

rolling updates as land-use applications are filed · public

facility_performanceEU data centres within the directive's reporting scope

EU Data Centre Energy Performance Reporting Database

Creates a mandatory, comparable facility-performance layer spanning electricity, water, efficiency, and waste-heat indicators.

European Commission · Reported energy and water footprints plus aggregated EU data-centre performance indicators.

Annual reporting under the Energy Efficiency Directive framework · EU reporting database and public aggregate dashboard

facility_performanceIreland grid-connected data centres

Central Statistics Office Ireland — Data Centres Metered Electricity

Rare national primary meter: 23% of Ireland metered electricity in 2025 — a buyer-side system-impact index, not a developer claim.

Central Statistics Office Ireland · Official metered electricity consumption of grid-connected data centres as a share of national demand.

Annual statistical release · Public statistical release pages

facility_performanceUnited States

LBNL / DOE U.S. Data Center Energy Use Report

Primary U.S. energy baseline: 176 TWh (2023) and 325–580 TWh (2028 range) — the denominator for national share claims.

Lawrence Berkeley National Laboratory / U.S. Department of Energy · U.S. data-center electricity use history and 2028 ranges; IT vs facility splits where published.

Periodic national assessment · Public DOE article and LBNL PDF

factory ordersUnited States

Census M3 — electrical equipment new and unfilled orders (NAICS 335)

Monthly Manufacturers' Shipments, Inventories and Orders survey with downloadable XLSX tables breaking out NAICS 335 electrical equipment: May 2026 advance shows new orders of $18.4B and unfilled orders of $54.6B (+5.9% YoY NSA).

U.S. Census Bureau · national, by NAICS industry group

monthly (advance ~4 weeks after month end) · public

forecast + connection reportingNew Zealand

Transpower Te Kanapu data-center insights + connection reporting

System-planner data-center capacity scenarios and connection diligence. The companion connection-reporting process records proponent, site, coordinates, capacity, projected demand, commissioning date, and project stage; detailed portal access is controlled.

Transpower New Zealand · National data-center scenarios and project connection reporting

Te Kanapu insights are point-in-time; customer reporting updates with material changes · controlled access

forecast + firm capacitySingapore

Singapore Demand & Supply Outlook / new generation RFP

Primary demand-and-supply planning and firm-generation procurement signal. EMA identifies data centers and semiconductors as high-demand drivers; it does not identify individual data-center projects.

Energy Market Authority (EMA) · National demand and firm-generation procurement

Annual outlook; RFP schedule and awards as released · public

forecast + large-load validationPJM — United States

PJM data-center forecast + load adjustment evidence

Current long-term forecast report, tables, data workbook, data-center accuracy report, and load-adjustment evidence. PJM vets large-load requests for commitments and double-counting; this is a forecast and validation source rather than a public queue.

PJM Interconnection · PJM zone / load-serving entity / 20-year demand forecast

Annual forecast with current data workbooks and stakeholder materials · public

forecast + system planAustralia

AEMO digital-demand integration + 2026 Integrated System Plan

Official system-planning entry point for data-center demand integration, connection processes, forecast development, and the 2026 ISP modelling/data workbooks. It is a planning source, not a public project queue.

Australian Energy Market Operator (AEMO) · National Electricity Market and Wholesale Electricity Market planning

2026 ISP published June 2026; planning cycle and supporting data updated regularly · public

grid_carbonU.S. electric power sector

EPA Emissions & Generation Resource Integrated Database (eGRID)

Supports location-based carbon, marginal procurement context, and grid-resource-mix comparisons without relying on vendor sustainability claims.

U.S. Environmental Protection Agency · Plant, balancing-authority, state, and grid-subregion generation, fuel mix, emissions, and emissions-rate data.

Annual edition with periodic revisions · Public workbooks, files, and explorer tools

grid_demandGlobal; scenario-based outlooks through 2030/2035

IEA Energy and AI — Data-centre electricity demand

Authoritative global demand envelope (base case ~945 TWh by 2030, just under 3% of world electricity) used to size national/grid conversations without confusing MW requests for metered load.

International Energy Agency · Global data-centre electricity consumption history and scenarios; AI-accelerated vs conventional server splits.

Major report cycles with interim updates · Public report pages and downloadable charts

grid_demandU.S. balancing authorities and planning areas subject to Form 714

FERC Form No. 714

Adds hourly and forecast load context for testing whether data-center requests are large relative to the system that must serve them.

Federal Energy Regulatory Commission · Balancing-authority hourly system demand, peak-load forecasts, planning areas, and transmission-planning information.

Annual filing · Public bulk data and filing downloads

grid_demandU.S. electric utilities and balancing authorities

EIA Form 861 Electricity Data

Provides the denominator and utility-level context needed to compare proposed hyperscale load with existing customers, sales, and programs.

U.S. Energy Information Administration · Utility service territory, sales, revenue, customers, demand response, distributed generation, and reliability records.

Annual, with early-release and final files · Public spreadsheets and EIA API products

grid_demandUnited States commercial building stock

EIA Annual Energy Outlook — commercial data-center servers

Places data-center servers inside the official long-run electricity outlook rather than as an orphan side estimate.

U.S. Energy Information Administration · Long-run U.S. commercial-sector server electricity projections, including standalone data centers.

Annual outlook · AEO browser tables and Today in Energy notes

grid_pipelineNational Electricity Market, with current concentration in New South Wales and Victoria

AEMO Data Centre Connection Pipeline

Offers a rare operator-published leading indicator: projects in application, approval, and implementation before they appear in metered demand.

Australian Energy Market Operator · Large data-centre connection projects by stage, aggregate maximum demand, and state concentration.

Quarterly Energy Dynamics reporting · Public report PDF and agency news summary

grid_pipelinePJM footprint

PJM Planning — data-center load growth outlook

Operator planning forecast (e.g. up to ~30 GW by 2030) for the densest U.S. interconnection market outside ERCOT deep-dive coverage.

PJM Interconnection · RTO planning statements on projected data-center load growth through the planning horizon.

Annual year-in-review and planning updates · Public PJM communications

historical load + generationUnited States

NREL ARPA-E PERFORM load, wind, and solar profiles

Time-coincident load, wind, and solar generation profiles. ERCOT actuals cover 2017-2018; NYISO/MISO/SPP actuals cover 2018-2019. Includes deterministic and probabilistic forecasts for existing sites and projected capacity-growth scenarios derived from interconnection-queue proposals.

National Renewable Energy Laboratory (NREL), via OpenEI · 5-minute resolution time series

Static historical archive (published 2022-08-18; last revised 2022-10-06) · public

interconnection queueUnited States

LBNL Queued Up: US interconnection queue dataset

The canonical US-wide interconnection-queue dataset covering all 7 ISOs/RTOs plus ~50 non-ISO utilities (~98% of US generating capacity), with project-level Excel data on capacity by fuel, status, region, queue duration and completion rates. The 2026 edition covers ~8,200 active projects with data through end-2025.

Lawrence Berkeley National Laboratory · project-level, by fuel, status, region, cohort year

annual (data file ~May, report ~June) · public

large-load operationsCentral United States

MISO large-load operational forecasting requirements

Operator requirements and planning context for non-conforming large loads, including data centers and advanced manufacturing. Useful for operational-readiness signals; it is not a public, project-level queue.

Midcontinent Independent System Operator (MISO) · Large-load operational forecasting requirements

Current stakeholder / operator guidance · public

large-load queueItaly

Terna Econnextion grid connection requests dashboard

Monthly dashboard of high-voltage connection requests to the Italian grid that explicitly breaks out data centers as a consumer-load category, with request counts, GW and authorisation status navigable to regional/provincial level. Bulk files available via the portal's Download Center.

Terna (Italian TSO) · request-level aggregates by category, region/province and status

monthly · public

large-load queueNew York (NYISO)

NYISO interconnection queue (incl. load facilities)

Public interconnection queue spreadsheet whose scope explicitly includes new and materially modified load facilities alongside generation and transmission, with queue position, MW, zone, status and dates per project.

New York ISO · project-level (MW, type incl. load, zone, status, dates)

continuously maintained spreadsheet · public

large-load queueNorway

Statnett capacity reservations + mature connection queue

Transmission-system connection data published from Statnett’s connection database: reserved capacity and mature queue projects by company, industry category, MW volume, location, and planned completion. Category filtering can surface data-center / industrial demand where disclosed.

Statnett · Project / customer / industry category / capacity reservation

Updated as Statnett releases connection-database statistics · public

large-load queueTexas — United States

ERCOT Large Load Interconnection (LLI) + LLWG

Project-scale large-load request workflow, study and fee milestones, energisation path, status materials, and committee records. This is the current deepest public data-center / large-load queue in the registry.

Electric Reliability Council of Texas (ERCOT) · Large-load applications and committee materials; Texas interconnection system

As released by ERCOT and the Large Load Working Group · public

large-load queue + adequacyIreland and Northern Ireland

EirGrid / SONI large-demand system evidence

Primary system-operator paper with operating data-center peak demand, contracted capacity, demand share, connection applications, technical constraints, and 2032 adequacy scenarios. Some figures combine data centers with other new-technology loads and are labelled that way in Grid Watch.

EirGrid / SONI · All-island system and transmission-demand evidence

Point-in-time paper (November 2025); follow EirGrid/SONI for revisions · public

market vacancyglobal (~16 major markets, North America detail)

CBRE Global and North America Data Center Trends

Free brokerage research with market-by-market vacancy, net absorption MW, under-construction MW and asking-rate trends; the 2026 edition reports North America top-4 net absorption of 2,236 MW in Q1 2026 and Northern Virginia vacancy of 0.3%.

CBRE Research · metro market level (vacancy %, absorption MW, supply MW)

annual global edition plus semiannual North America edition · public summary

memory disclosureUnited States / global

Micron quarterly results via SEC EDGAR

Quarterly earnings 8-Ks on EDGAR with detailed HBM product and supply disclosures; FQ3-2026 revenue was $41.46B vs $9.30B a year earlier, with HBM4 in high-volume shipment and FQ4 guidance of ~$50B. EDGAR is the bot-friendly path (Micron's IR site times out).

Micron Technology / U.S. SEC EDGAR · company, quarterly

quarterly (fiscal quarters end Feb/May/Aug/Nov) · public

memory spot pricesglobal

DRAMeXchange spot price board

Daily DRAM and NAND flash spot prices by part (DDR5/DDR4 by density, NAND) plus the DXI market index, updated intraday; DDR5 16Gb averaged $47.80 on July 10, 2026.

TrendForce (DRAMeXchange) · per-component spot price

daily (intraday updates) · public summary

networkENERGY STAR-certified large network equipment

ENERGY STAR Certified Large Network Equipment

Adds the often-missed network portion of IT load and makes RFP shortlists auditable at model level.

U.S. Environmental Protection Agency · Switch/router product configuration, throughput, port, power, and efficiency fields.

Continuously maintained product registry · Socrata API and CSV export

network infrastructureworldwide

TeleGeography Submarine Cable Map GeoJSON API

The industry-standard submarine cable map exposes free unauthenticated GeoJSON endpoints for cable routes and landing points (695 cable systems and 1,917 landing points verified on 2026-07-12), with per-cable owner, length and ready-for-service metadata.

TeleGeography · per-cable-system geometry and metadata; per-landing-point coordinates

continuously updated (snapshot diffing yields announcements series) · public API

nuclear ppaUnited States (PJM, MISO)

Constellation Energy newsroom — data-center nuclear PPAs

Primary press releases for the flagship operating-fleet nuclear PPAs with data-center buyers: Microsoft-Crane Clean Energy Center (~835 MW Three Mile Island restart, 20-year) and Meta-Clinton (1,121 MW, 20-year from June 2027), with MW and duration on-page.

Constellation Energy Corporation · per-deal (plant, MW, contract years, start date)

event-driven press releases · public

ocean computeChina

Highlander / HiCloud undersea data-center platform

Vendor and deployment source for underwater intelligent computing cabins, private cloud, undersea data centers, deployment projects, and the 2026 marine computing-power / offshore-energy partnership.

Beijing Highlander Digital Technology Co., Ltd. · Undersea data-center product, deployment, and project evidence

Company updates and project announcements · public

ocean computePacific Ocean and North Sea

Microsoft Project Natick — subsea data-center research record

Primary mission record for the two underwater prototypes: deployment, retrieval, power, rack/server payload, storage, reliability, water use, and the current research-stage status.

Microsoft Research · Prototype deployment and mission evidence; Pacific and North Sea

Completed pilot; mission record retained for design and reliability evidence · public

ocean computeUnited States / United Kingdom / Norway

Subsea Cloud Project OTTO

Commercial subsea data-center platform describing modular build, deployment, maintenance, cooling, connectivity, and public performance claims. No public MW or operating customer count is assumed.

Subsea Cloud Inc. · Subsea modular data-center product and planned deployments

Company material; project status must be verified directly · public

oem backlogglobal (US-heavy demand)

Siemens Energy quarterly earnings — orders and backlog

Quarterly disclosure of orders (record €17.7B in Q2 FY2026), order backlog (€154B) and book-to-bill (1.72) by segment; Grid Technologies covers HVDC, switchgear and grid-connection equipment with US-driven growth.

Siemens Energy AG · company-level by segment

quarterly (fiscal year ends September) · public

oem backlogglobal (US-heavy)

GE Vernova quarterly earnings — gas turbine GW slot reservations

Quarterly disclosure of total orders ($18.3B in Q1 2026), total backlog ($163B) and, uniquely, gas power equipment backlog plus slot-reservation agreements in GW (83 to 100 GW during Q1 2026, guiding to 110+ GW by end-2026).

GE Vernova · company-level by segment; gas backlog in GW

quarterly · public

operating fleetUnited States

EIA Form 860 — Annual Electric Generator Inventory

Generator-level inventory: nameplate capacity, fuel type, owner/utility, in-service and planned/actual retirement dates, prime mover, technology, and environmental equipment. Texas/ERCOT plants are a filterable state subset, not a separate file.

U.S. Energy Information Administration (EIA), U.S. Department of Energy · Individual generator/unit level; United States

Annual. 2025 early-release data published 2026-06-09; final 2025 data due September 2026. · public

opposition trackerUnited States

Data Center Watch quarterly opposition report (10a Labs)

Quarterly tracker of US data-center projects blocked or delayed by local opposition, moratorium proposals and state legislation; Q1 2026 counted at least 75 projects worth ~$130B blocked or delayed, with 300+ state bills filed in six weeks. Note the name collision with datacentr.net's own branding — attribute carefully.

Data Center Watch / 10a Labs · project-level and state-level rollups

quarterly · public

orbital computeLEO / ISS

Axiom Space Orbital Data Centers

Primary orbital data-center program record covering AxDCU-1, the first two free-flying ODC nodes launched January 11, 2026, optical links, the ISS node roadmap, and long-term kilowatt-to-megawatt expansion intent.

Axiom Space · Orbital node / ISS prototype / network roadmap

Company program page updated as missions and nodes progress · public

physical_riskUnited States, county and Census-tract geographies

FEMA National Risk Index Data

Adds a consistent multi-hazard screen for flood, hurricane, wildfire, heat, winter weather, earthquake, and other correlated siting risks.

Federal Emergency Management Agency · County and Census-tract hazard exposure, expected annual loss, social vulnerability, and community resilience measures.

Periodic methodology and data releases · Public CSV/GIS downloads and map

powerENERGY STAR-certified UPS products

ENERGY STAR Certified Uninterruptible Power Supplies

Supports primary/secondary/tertiary power-chain scoping with comparable UPS ratings and part-load efficiency.

U.S. Environmental Protection Agency · UPS topology, rated output, input/output voltage, load-dependent efficiency, and product status.

Continuously maintained product registry · Socrata API and CSV export

powerU.S. liquid-immersed, low-voltage dry-type, and medium-voltage dry-type distribution transformers

DOE Distribution Transformer Standards and Test Procedures

Anchors transformer efficiency, compliance, and replacement assumptions while supplier lead-time signals are assessed elsewhere.

U.S. Department of Energy · Federal definitions, test procedures, conservation standards, and rulemaking evidence for distribution transformers.

Rulemaking and standards cycles · Public rules, notices, and supporting documents

ppa pricesNorth America (separate Europe edition)

LevelTen PPA Price Index quarterly summaries

Quarterly P25 solar/wind PPA offer-price index built from marketplace offers across AESO, CAISO, ERCOT, MISO, PJM and SPP; public posts carry directional QoQ/YoY changes each quarter (Q1 2026: P25 solar +4.6% QoQ, wind +8% QoQ), with absolute levels subscriber-only.

LevelTen Energy · continental index plus per-ISO commentary, P25 offer prices

quarterly · public summary

procurementTenders Electronic Daily notices across the EU and associated jurisdictions

TED Search API

Provides an open, cross-border buyer-intent feed for data-centre construction and equipment with both pre-award and award-stage evidence.

Publications Office of the European Union · EU procurement notices with buyers, CPV classifications, values, procedure stages, places of performance, deadlines, awards, and notice text.

Near-daily notice publication · Open Search API; no authentication required

procurementU.S. federal contract opportunities

SAM.gov Get Opportunities Public API

Surfaces buying intent and RFP language for servers, accelerators, cooling, UPS, generators, substations, construction, and operations before awards.

U.S. General Services Administration · Active and archived U.S. federal solicitations, notices, agencies, NAICS/PSC codes, response dates, place of performance, and attachments metadata.

Active opportunities daily; archived opportunities weekly · API key required; public API

producer pricesUnited States

BLS Producer Price Index API — grid equipment and electrical series

Official monthly producer price indexes via free JSON API for transformers (PCU335311335311), switchgear/switchboards (PCU335313335313) and electrical contractors on nonresidential work (PCU23821X23821X). Transformer PPI is up ~5.5% YoY as of May 2026 and has more than doubled since pre-2020.

U.S. Bureau of Labor Statistics · national monthly index by NAICS industry

monthly (mid-month, one-month lag) · public API

producer_indexU.S. producer prices for multiuser host computers

BLS Producer Price Index — Host Computers / Servers (WPU11510116)

Producer-side price index for server hardware used in RFP escalation clauses and CapEx sensitivity screens.

U.S. Bureau of Labor Statistics · Monthly producer prices for host computers / multiuser servers (mainframes, Unix, and PC servers).

Monthly · BLS PPI detailed tables and FRED series WPU11510116

producer_indexU.S. semiconductor and electronic-component producers

BLS Producer Price Index — Semiconductor & Other Electronic Components (PCU3344133441)

Upstream silicon/component price pressure that eventually flows into accelerator boards and networking silicon.

U.S. Bureau of Labor Statistics · Monthly producer prices for semiconductor and other electronic component manufacturing.

Monthly · FRED series PCU3344133441 (BLS series ID PCU33441-33441-)

project_pipelineTexas projects participating in the qualifying data-centre sales-tax exemption

Texas Qualifying Data Centers Registry

A direct project and customer-identity signal that appears earlier than many commercial inventory datasets and can be reconciled to permits and grid requests.

Texas Comptroller of Public Accounts · Registered qualifying data centres with owner, operator, occupant, certification/effective dates, and exemption status.

Administrative updates as projects qualify or status changes · Public state registry/list downloads

regulatory docketUnited States (federal)

FERC large-load co-location dockets (AD24-11 / EL25-49)

FERC's generic proceeding on large loads co-located at generating facilities (AD24-11) plus the related company-specific dockets (Talen/AWS ISA amendment, BGE/PECO/ComEd/Pepco filings) and the EL25-49 show-cause proceeding on PJM co-location rules. Full filing history available via FERC eLibrary docket sheets.

Federal Energy Regulatory Commission · filing-level (orders, comments, transcripts) per docket

event-driven as dockets progress · public

reliability + adequacyNorth America

NERC Long-Term Reliability Assessment

North American reliability assessment with planner-submitted demand forecasts, large-load / data-center growth, resource-adequacy risk, and regional exposure. The source is a validated planning outlook, not an interconnection queue.

North American Electric Reliability Corporation (NERC) · Assessment area and regional planning outlook

Annual assessment · public

retail pricesUnited States (50 states + DC)

EIA Electric Power Monthly — industrial retail price by state

Monthly average retail electricity price by state and sector with YoY comparison (Table 5.6.A); April 2026 shows US industrial average 8.66 c/kWh, Texas 6.33, Virginia 9.86. Also queryable via the EIA v2 API for automation.

U.S. Energy Information Administration · state by sector, monthly

monthly · public

semiconductor billingsglobal (4 regions)

WSTS historical semiconductor billings dataset

Free downloadable Excel of monthly worldwide semiconductor billings by region spanning four decades, updated monthly from the WSTS Blue Book; latest data May 2026. This is the machine-readable series behind SIA's monthly press numbers ($120.6B in May 2026, +104.1% YoY).

World Semiconductor Trade Statistics · region, monthly, machine-readable

monthly · public

semiconductor revenueTaiwan / global

TSMC monthly revenue disclosure

Statutory monthly revenue disclosure (NT$ millions, MoM/YoY, YTD) published around the 10th of each month; May 2026 revenue was up 30.1% YoY. TSMC fabricates nearly all NVIDIA/AMD accelerator dies.

Taiwan Semiconductor Manufacturing Company · company, monthly

monthly (~10th of following month) · public

storageENERGY STAR-certified block-I/O data-center storage products

ENERGY STAR Certified Data Center Storage

Connects storage architecture and capacity procurement to measured energy characteristics rather than nameplate guesses.

U.S. Environmental Protection Agency · Block-storage model capacity, controller, drive, performance, and energy-efficiency attributes.

Continuously maintained product registry · Socrata API and CSV export

supplier_demandTWSE-listed companies, including major ODM and semiconductor suppliers

TWSE OpenAPI

Creates a high-frequency producer-side view of AI server, PCB, power, cooling, and semiconductor demand before annual financial statements.

Taiwan Stock Exchange · Listed-company monthly revenue, filings, securities, and market datasets for Taiwan's server and semiconductor supply chain.

Daily market data and monthly/periodic company filings · Public OpenAPI endpoints

system planJapan

OCCTO electricity supply plan aggregation (Japan)

Japan's cross-regional grid coordinator publishes English editions of the annual Aggregation of Electricity Supply Plans and seasonal supply-demand outlooks, with 10-year demand and capacity projections for the 10 service areas; data-center and chip-fab demand upgrades surface here first.

Organization for Cross-regional Coordination of Transmission Operators, Japan · national and 10 service-area aggregates, 10-year projections

annual supply plan, seasonal outlooks · public

tax abatementsUnited States (state + local)

Good Jobs First Tax Break Tracker (GASB 77)

Searchable database of GASB 77 tax-abatement disclosures from state and local government financial statements, FY2015-FY2025, covering all 50 states plus major localities; a companion April 2026 report quantifies data-center-specific losses (Georgia $2.5B, Virginia $1.94B, Texas $1B per year).

Good Jobs First · jurisdiction by program by fiscal year

rolling updates as financial reports are filed · public

tradeOfficial U.S. merchandise trade statistics

USITC DataWeb Query API

Measures physical supply flow and price/value pressure for transformers, switchgear, generators, chillers, servers, semiconductors, and optical equipment.

U.S. International Trade Commission · Commodity-level U.S. imports/exports by HTS/SITC/NAICS, partner, customs district, quantity, value, and tariff program.

Monthly trade-data releases and revisions · Free registration and API key required

transformer supplyUnited States

Wood Mackenzie North America transformer market releases

Free press-release summaries of Wood Mackenzie's paid transformer supply and lead-time survey: 30% power-transformer and 10% distribution-transformer supply deficits in 2025, demand +116% since 2019, imports covering 80% of power-transformer supply, and average lead times around 120 weeks.

Wood Mackenzie · US market aggregate, by transformer class

irregular (roughly quarterly-to-annual public updates) · public summary

underground computeItaly

Intacture active-mine data center

Company announcement for an operating data center approximately 100 meters below the surface inside an active mine in Val di Non.

GPI Group · Facility / operating launch

Operating from June 2026 · public

underground computeNorway

Green Mountain SVG-Rennesøy mountain-hall data center

Operating high-security data center in converted mountain halls / former NATO ammunition storage, with published site capacity, space, PUE, grid supplies, and build-out description.

Green Mountain · Facility site and capacity; Stavanger region

Operating facility; site updates as expansion changes · public

underground computeNorway

Lefdal Mine Data Centers facility

Operating underground data-center campus in a former mine with published chambers, white-space potential, levels, power backbone, and 80 MW Level 3 capacity.

Lefdal Mine Data Centers · Facility / chamber / level buildout

Natural-mine campus with staged capacity expansion · public

underground computeSweden

Bahnhof Pionen underground data center

Operating Stockholm data center built inside a former Cold War civil-defense / nuclear fallout shelter, with published security, connectivity, cooling, and operations detail.

Bahnhof · Facility and colocation infrastructure

Operating facility; capacity not publicly stated on the source page · public

underground computeSwitzerland

Swiss Fort Knox underground data center

Operating secure underground data-storage / data-center infrastructure in former Swiss military bunkers in the Alps.

MOUNT10 AG · Facility and secure-storage infrastructure

Operating facility; capacity not assumed where not published · public

underground computeUnited States

Iron Mountain Western Pennsylvania WPA-1

Operating colocation site 220 feet underground in a former limestone mine, with published 15.5 MW power capacity, 330,000 square feet, geothermal cooling, utility, and connectivity.

Iron Mountain Data Centers · Facility-level capacity and infrastructure

Operating facility; published capacity subject to commercial availability · public

waterUnited States monitoring network

USGS Water Data APIs

Makes watershed availability, drought response, intake/discharge context, and cooling-water risk measurable at site scale.

U.S. Geological Survey · Real-time and historical streamflow, groundwater, water-quality, daily-value, and monitoring-location records.

Real-time to daily, depending on station and parameter · Modern public REST APIs and downloadable data

waterUnited States; highly local

U.S. Army Corps / state water-withdrawal permit portals

Cooling topology is often decided by water rights and discharge permits before a CDU model is selected.

State water boards and U.S. Army Corps of Engineers (jurisdiction-dependent) · Intake, discharge, and water-rights dockets that constrain evaporative and once-through cooling designs.

Project-driven administrative dockets · State portals and federal permit systems; coverage fragmented

water disclosureglobal (Google owned/operated data centers)

Google Environmental Report (data center water, site-level)

Annual environmental report disclosing total operational water (2024: 11,011 million gallons withdrawn, 8,135 consumed, +28% YoY), 64% freshwater replenishment, and per-site water data across Google's data center cities.

Google / Alphabet · company-wide totals plus site/city-level data center water figures

annual (mid-year) · public

water stressUnited States

US Drought Monitor data downloads

Weekly national drought classification (D0-D4) with downloadable statistics by state, county and urban area, DSCI index, GIS shapefiles and web services for programmatic pulls.

National Drought Mitigation Center (Univ. of Nebraska-Lincoln) with USDA and NOAA · county (and finer GIS), state, national

weekly (Thursdays) · public API

wholesale pricesUnited States (7 major hubs)

EIA Wholesale Electricity Market Data (ICE day-ahead hub prices)

Free republication of ICE-traded daily day-ahead firm power prices (weighted-average index, high/low, volume) at seven hubs including ERCOT North and PJM West, with current-year and multi-year XLSX archives.

U.S. Energy Information Administration · daily, per trading hub, price and volume

biweekly updates of daily data · public

wue disclosureglobal (Meta owned data centers)

Meta Environmental Data Index (fleet WUE/PUE and water tables)

Machine-readable appendix to Meta's annual sustainability report with 2020-2024 time series for data-center WUE (0.30 down to 0.19 L/kWh), PUE (1.10 to 1.08), water withdrawal by source in megaliters, and withdrawals from high-water-stress areas.

Meta · fleet-wide annual series, split by water source and stress class

annual (~August, with sustainability report) · public

wue disclosureglobal (Microsoft owned fleet)

Microsoft datacenter efficiency disclosures (PUE/WUE by region)

Official fleet efficiency page publishing PUE and WUE for Microsoft's owned datacenter fleet, globally and by Americas/EMEA/APAC, for the last two fiscal years: FY25 global WUE 0.27 L/kWh and PUE 1.17.

Microsoft · fleet-wide plus 3 regions

annual (fiscal-year update) · public

Data center jobs direct employer · sponsored backfill

Critical-facilities, MEP, commissioning, and operations roles from public Greenhouse, Lever, and Ashby boards — plus optional sponsored listings when Adzuna keys are configured. No Indeed or LinkedIn scraping.

Email when new data-center roles are posted. Alerts cover direct employer rows only. Double opt-in — confirm via the email we send.

Browse 171 open roles · JSON · CSV

Want filtered alerts by county and role type? Pro database ($25/mo when live) bundles project-level LLI tracking with job alerts. Free digest: weekly signal.

Prepare for these roles

Certification and resume tools common in data-center hiring pipelines.

Some apply links are sponsored listings served via Adzuna; Data Center Watch may earn a fee if you click or apply. Rows marked Direct employer link straight to company ATS career pages.

Some apply links are sponsored listings served via Adzuna; Data Center Watch may earn a fee if you click or apply. Rows marked Direct employer link straight to company ATS career pages.

Get the signal

Daily

Daily digest signup not configured yet — see deploy/daily-digest-setup.md.

Weekly

Sourced links only, no fluff. Double opt-in — check your inbox (and spam) for the confirmation link.

Pro database

Project-level LLI tracking, county permits, CSV/API export, docket alerts.

Open pro database Free while in beta

Browse the full database · JSON API · CSV export

Live activity console automation, subscribers & new postings · polled ~60s

A public, no-PII feed of what the automation behind Data Center Watch is doing right now — pipeline refresh runs, curator/audit agent outcomes, new job postings, and anonymized subscriber signups. This never shows an email address.

gridwatch@vps:~$ tail -f activity.json
connecting…

Frequently asked questions

What is Data Center Watch?

A primary-source intelligence dashboard for data-center power: interconnection queues, live grid demand, generation capacity, HPC systems, and the power, cooling, and supply-chain dependencies behind new sites.

Where does the data come from?

Every figure links to its primary source. Live grid demand and generation capacity come from EIA Forms EIA-930 and EIA-860/860M. ERCOT queue and watch-feed figures come from its own LLI status updates, TAC/LLWG decks, GIS reports, and PUCT filings. The remaining sources are catalogued in the worldwide connection & buildout sources section — nothing is interpolated to fill a gap.

What does the ERCOT deep dive mean?

ERCOT has the most complete public interconnection-queue reporting, so Data Center Watch tracks its project-level queue and watch-feed line items in the greatest depth. Other regions get live demand and generation capacity from EIA but not a project-level queue.

Is Data Center Watch affiliated with ERCOT, EIA, or any grid operator?

No. Data Center Watch is not affiliated with ERCOT, EIA, or any grid operator. Data may contain errors — verify against primary sources before making decisions.

Is there a free API or data export?

Yes. The facility registry is downloadable as JSON and CSV, the news monitor is available as RSS, and the pro database has its own JSON API — see the worldwide connection & buildout sources section and the pro database section for links.

How often is the data updated?

An automated pipeline refreshes the underlying data multiple times a day; the live activity console on this page shows each refresh run as it happens.