Where AI Data Centers Are Located in Washington State

Introduction: Washington's explosive growth as an AI data-center hub

Washington State has quietly become one of the most strategically important AI infrastructure corridors in the United States. With abundant hydropower, favorable tax conditions, and a mature fiber network, the state now hosts over 113 operating or listed data-center sites, with at least 8 more under active construction, according to FP Analytics (2025).

At Pickastor, our analysis of cloud AI deployment patterns shows that e-commerce businesses, from independent sellers to enterprise retail teams, are increasingly dependent on the very infrastructure concentrated in Washington's river valleys and suburban corridors. When that infrastructure faces capacity pressure, the downstream effects on AI-powered product optimization, search ranking, and inference latency are real and measurable.

A power grid under pressure

The scale of demand is striking. Data centers already account for approximately 5.7% of Washington's total electricity consumption, a figure that could climb to roughly 10% of statewide electricity use under a high-growth scenario, according to FP Analytics (2025). In Central Washington, where hydropower is cheapest and most accessible, vacancy rates are approaching zero.

Why this matters for e-commerce teams

For SMB owners and enterprise e-commerce teams relying on cloud AI services, understanding where this infrastructure sits, and where it is constrained, is no longer an academic exercise. It shapes vendor selection, latency expectations, and long-term platform resilience. This study maps exactly that landscape.

Methodology: how we sourced and verified this data

This study draws exclusively on published, verifiable research from recognized industry and policy authorities. Every figure cited in the sections that follow has a named source, a publication year, and a direct link for independent verification.

Primary sources

The core data in this study comes from three authoritative bodies:

• FP Analytics (2025): Electricity demand modeling and AI infrastructure capacity projections, drawn from their report Powering the AI Era
• CBRE (2025): Commercial real-estate absorption rates, vacancy data, and regional market sizing for Pacific Northwest data-center markets
• Northwest Power and Conservation Council: Utility load forecasting and grid-capacity planning documents covering Washington's hydropower-dependent regions

Data freshness and verification standards

2024 and 2025 figures are prioritized throughout. Where older benchmarks appear, they serve only as historical reference points, clearly labeled as such. No estimates, projections, or unattributed figures are included.

Regional breakdowns are cross-referenced against both commercial real-estate reports and utility planning filings to reduce single-source bias. Readers who want broader context on how raw data feeds AI systems can consult Everything You Need to Know About Data for AI.

Washington's data-center footprint: 113 operating sites and growing

Washington State currently hosts 113 operating data centers, with at least 8 additional sites under active construction. That footprint places the state among the most significant infrastructure markets in the country, and the numbers are still moving upward as hyperscale operators commit to larger, more power-intensive campuses built specifically for AI and high-performance computing workloads.

Scale of electricity demand

According to FP Analytics (2025), Washington's data centers collectively consume 5.17 TWh of electricity annually, ranking the state ninth among all U.S. data-center markets by power demand. To put that figure in context, those facilities now account for approximately 5.7% of Washington's total statewide electricity consumption. A single sector drawing nearly one in every eighteen units of power generated across the state signals a structural shift, not a temporary spike.

Where growth is concentrated

Expansion is not happening evenly across the state. The Seattle metro area retains an established cluster of facilities, but the most aggressive growth is occurring outside the city. Central Washington, the Tri-Cities region, and several Eastern Washington counties are absorbing the bulk of new construction activity, driven largely by access to affordable hydroelectric power and available land.

This geographic spread matters for understanding how AI infrastructure is reshaping regional economies and utility planning. The push toward mega-campus formats, where a single operator may control hundreds of megawatts on one site, reflects the same underlying dynamic that has AI systems demanding ever-larger volumes of training data to sustain performance gains.

The next section examines Central Washington specifically, where vacancy rates have collapsed and demand has tripled in a compressed timeframe.

Central Washington's AI boom: near-zero vacancy and tripled demand

Central Washington has become the most capacity-constrained data-center market in North America. According to Cascade PBS (2026), the region's vacancy rate collapsed to just 0.16% in H2 2024, the lowest figure recorded across 18 North American markets tracked by CBRE. At that level, available space is effectively nonexistent for operators seeking immediate deployment.

Demand that tripled in a single year

The vacancy figure does not exist in isolation. Net absorption in Central Washington jumped from 20.9 MW in 2023 to 60 MW in 2024, a near-tripling driven by AI workload migration and aggressive hyperscaler expansion. That rate of uptake is extraordinary by any historical benchmark, and it compresses the timeline for operators hoping to secure capacity before lease rates climb further. The same forces reshaping how businesses think about product feed optimization for AI are, at a structural level, driving the compute demand that has emptied Central Washington's available inventory.

Seattle's parallel tightening

Seattle's market is following a similar, if less extreme, trajectory. CBRE data for H2 2024 recorded 11.1 MW of net absorption in the Seattle market alongside 8.5 MW under active construction. The supply-demand imbalance across both markets is pushing lease rates upward and drawing new operator investment into the state, a dynamic the next section examines through the lens of power availability.

Power constraints: the real bottleneck for AI infrastructure expansion

Electricity availability has overtaken land and labor as the primary constraint on data center growth in Washington State. New facility approvals, expansion timelines, and site selection decisions are now shaped almost entirely by what the grid can deliver, not what developers can build.

How large could the electricity demand become?

The scale of projected demand is striking. According to FP Analytics (2025), Washington's data center electricity consumption could reach roughly 10% of total statewide electricity use under a high-growth AI scenario within five years. The Northwest Power and Conservation Council projects an additional 2.4 to 4 GW of regional electricity load from data centers and chip fabrication facilities by 2029, a figure that puts serious pressure on existing generation and transmission capacity.

Why hydropower is no longer a guaranteed advantage

Washington's hydropower network has long been a competitive asset for data center operators. That advantage is narrowing. Current transmission infrastructure was not designed to route power at the volumes AI workloads require, and utilities are now prioritizing grid upgrades and new generation investment to keep pace with demand. Interconnection queues are lengthening, and some projects face multi-year waits for grid access.

Understanding how these infrastructure dynamics shape workforce and operational planning is worth exploring further, particularly for teams tracking how data-intensive roles are evolving alongside AI adoption.

Power availability is now the variable that determines whether a Washington data center project moves forward at all.

Year-over-year trends: acceleration in 2024 and 2025

The numbers tell a clear story: Washington State's data center market did not grow incrementally in 2024, it accelerated sharply. Central Washington net absorption tripled year-over-year, and construction pipelines expanded in parallel, signaling that operators are committing capital well ahead of current demand curves.

Central Washington's absorption surge

According to CBRE (2025), Central Washington recorded 60 MW of net absorption in 2024, compared to just 20.9 MW in 2023. That near-tripling in a single year reflects the consolidation of AI workloads into a region that offers hydroelectric power, cooler ambient temperatures, and relatively affordable land. The shift is not a one-quarter anomaly. It represents a structural reorientation of AI infrastructure investment toward the Columbia Basin.

Seattle's construction pipeline holds firm

Seattle posted 11.1 MW of net absorption in H2 2024 alongside 8.5 MW under construction during the same period, per CBRE (2025). That simultaneous expansion of both occupancy and new supply indicates sustained operator confidence rather than speculative overbuilding.

Emerging mega-campus proposals and rising scrutiny

Proposed campuses across Grant, Douglas, Chelan, Benton, and Walla Walla counties are substantially larger than any historical Washington facilities, with some exceeding 1,000 MW in planned capacity. According to MRSC (2026), local governments are responding with increasing scrutiny around water usage, emissions, tax exemptions, and infrastructure strain. Lease rates are rising as vacancy tightens, though Washington's overall cost-of-occupancy position remains competitive relative to other major U.S. markets. For teams evaluating AI infrastructure dependencies, including those working with AI data labeling companies, understanding where capacity is concentrating matters for vendor selection and latency planning.

Regional breakdown: where AI data centers are concentrated in Washington

Washington's data center capacity is not evenly distributed. Three distinct geographic corridors have emerged, each shaped by different combinations of power access, real estate economics, and proximity to existing infrastructure. Understanding where capacity is concentrating helps organizations make more informed decisions about vendor selection and latency planning.

Central Washington: the primary hyperscale hub

Grant, Douglas, and Chelan counties have become the dominant destination for new large-scale investment. According to CBRE (2025), vacancy in Central Washington sits at just 0.16%, with 60 MW of net absorption recorded in 2024 alone. Direct access to Columbia River hydropower, combined with lower real estate costs than Seattle, makes this corridor the preferred location for hyperscale campuses requiring raw power at scale.

Seattle: established infrastructure, slower growth

Seattle remains a significant secondary market, anchored by major cloud provider headquarters and a deep talent pool. However, power constraints and higher land costs have slowed new development relative to Central Washington. Growth here tends to favor colocation expansions and enterprise deployments rather than greenfield hyperscale builds.

Tri-Cities: the emerging third corridor

Benton, Franklin, and Walla Walla counties are attracting attention from developers seeking room to build at scale. Several large proposed campuses are currently under development in the region, positioning the Tri-Cities as a meaningful expansion corridor over the next several years. For e-commerce teams evaluating AI tooling performance and infrastructure dependencies, including those using platforms like WooCommerce AI optimization solutions, the geographic distribution of this capacity has direct implications for latency and provider reliability.

Key takeaways: what the data tells us about Washington's AI infrastructure future

The numbers tell a clear story: Washington State has moved well beyond regional significance to become a structural pillar of the United States AI infrastructure landscape. The data points below synthesize what the evidence means for businesses that depend on cloud AI services.

Washington is a top-tier market, and the numbers confirm it

With 113 operating or listed facilities and at least 8 more under active construction, according to FP Analytics (2025), Washington's data center footprint is not a regional curiosity. It is a nationally significant concentration of compute capacity, anchored by hyperscalers and cloud providers whose infrastructure directly underpins the AI tools that e-commerce businesses use daily.

Power is the binding constraint on future growth

Hydropower and tax exemptions have historically made Washington one of the lowest-cost operating environments for data centers. That cost advantage remains real, but electricity availability is now the limiting factor. AI workloads already account for 5.7% of Washington's total electricity demand, and under a high-growth scenario that figure could approach 10% of statewide consumption. Supply is being tested.

Central Washington's near-zero vacancy signals structural demand pressure

A 0.16% vacancy rate in Central Washington is not a market in equilibrium. It reflects AI workload demand that is outpacing available capacity, compressing timelines for new builds and applying upward pressure on pricing. Proposed mega-campuses represent a direct response to this imbalance, signaling a shift toward larger, AI-optimized facilities.

What this means for e-commerce teams

For businesses relying on cloud AI services, including generative search, product optimization, and recommendation engines, Washington's infrastructure trajectory points toward sustained, long-term capacity availability. Understanding why many e-commerce stores remain invisible to AI-driven discovery starts with recognizing that the infrastructure to support those systems is expanding rapidly. The foundation is being built. The question is whether your products are positioned to benefit from it.

Frequently asked questions

Are there AI data centers in Washington State?

Yes. According to FP Analytics (2025), Washington had 113 data centers as of May 2025, with at least eight additional sites under construction. The state is one of the most active AI infrastructure markets in North America.

Where are data centers located in Washington State?

Data centers are concentrated in two primary regions: the Seattle metro area on the west side of the Cascades, and Central Washington, particularly around Quincy and East Wenatchee. Central Washington has attracted significant investment due to its access to low-cost hydroelectric power.

What companies have data centers in Washington State?

Microsoft, Amazon Web Services, Google, and Meta all operate major facilities in Washington. Microsoft has been particularly active, with ongoing AI-focused expansion across multiple sites in the state.

Why are data centers being built in Central Washington?

Central Washington offers abundant, affordable hydropower and large parcels of land. Research suggests vacancy rates in the region reached historic lows in 2024, reflecting intense demand from hyperscale operators.

How many data centers are in Washington State?

Washington had 113 listed data center sites as of May 2025, with more under active development.

Is Microsoft building AI data centers in Washington State?

Yes. Microsoft has publicly committed to expanding its AI infrastructure footprint in Washington as part of broader national investment plans.

Does Washington State have enough power for AI data centers?

Power availability is an active concern. The Washington Data Center Workgroup has noted that "data centers are the largest source of expected electricity-load growth in the Pacific Northwest and could require substantial expansion of generation, substations, and transmission capacity."

What is the largest data center in Washington State?

Microsoft and Amazon operate some of the largest individual campus facilities in the state, though precise rankings by square footage shift as new phases come online.

Based on our work at Pickastor, businesses that understand the infrastructure powering AI discovery are better positioned to act on it. The Pickastor AI Optimization Platform helps e-commerce teams ensure their products are structured and visible within the AI-driven systems that Washington's expanding data center capacity supports.