CoreSite | AI Transformation Index
CoreSite is one of the cleanest AI-durable assets in the technology infrastructure universe — its core product (physical space, power, cooling, interconnec...
Overview
CoreSite operates multi-tenant colocation data center campuses, generating revenue by leasing physical rack/cage/suite space, high-density power and cooling, and network interconnection services (cross-connects, cloud onramps) to enterprises, cloud providers, and carriers on recurring lease-style contracts.
Executive Summary
CoreSite is one of the cleanest AI-durable assets in the technology infrastructure universe — its core product (physical space, power, cooling, interconnection) is not only impossible for AI to substitute but is actively demanded by every AI workload in the market, making AI a structural demand tailwind rather than a disruption threat. The internal AI transformation opportunity is limited because the production function is overwhelmingly physical, which means there is no large cost base of cognitive labor to compress. For a PE investor, this is a high-durability, low-disruption-risk infrastructure asset whose biggest AI-related risk is hyperscaler internalization of enterprise compute, not AI substitution of the colocation function.
Durability
How resilient is this business against AI disruption?
Market Substitutability
Profit Pool Migration
Barrier to Entry
Disintermediation Exposure
Switching Costs
Opportunity
How big a deal is AI for this business?
AI Applicability
Improvement Magnitude
Transformation Upside
Domain Intelligence Amplification
45.0
Ciridae AI Score
Durability Rationale
CoreSite's durability is structurally anchored in the physical irreplaceability of its assets: compute workloads — including AI workloads — must physically reside somewhere with regulated power, cooling, and low-latency fiber, and CoreSite owns those facilities in constrained Tier 1 metros. No amount of AI capability substitutes for the physical layer, and in fact the AI era is a demand accelerant. The combination of capex barriers to new entry, multi-year tenant commitments, meaningful switching costs, and growing interconnection network effects makes this one of the more structurally protected business models in the technology infrastructure landscape.
Opportunity Rationale
AI is not a large deal for CoreSite's internal operations — the business is fundamentally a physical asset operation where the cognitive labor component is relatively small, and the step-change AI can deliver to that labor is modest at the overall business level. Ironically, CoreSite's opportunity story runs the other direction: AI is a massive demand driver for its customers, accelerating consumption of the very physical infrastructure CoreSite provides, but that is a market tailwind rather than an internal transformation opportunity captured through AI applicability or magnitude.
Key Risks
Hyperscaler internalization risk is the primary durability threat: if enterprises accelerate full migration to public cloud (AWS, Azure, GCP) — potentially aided by AI reducing complexity of cloud migration — enterprise colocation demand could soften and CoreSite's interconnection moat weakens as on-premises footprints shrink.
The opportunity assessment could be wrong if AI-driven automation of facility operations (autonomous power management, predictive cooling, AI-orchestrated provisioning) proves far more impactful than expected — particularly if new AI-native competitors build next-generation facilities with dramatically lower opex structures, compressing incumbents' cost advantages.
Company Overview
| Products / Services | Colocation (cabinets, cages, private suites), interconnection and cross-connects, cloud onramps (native cloud onramps and Direct Connect/Dedicated Connect), software-defined networking (Open Cloud Exchange®), remote hands and professional services, AI-ready high-density rack and suite deployments |
AI & Tech Initiatives
| Name | Description |
|---|---|
| AI infrastructure solutions for enterprise AI workloads | CoreSite has explicitly positioned its colocation, high-density power and cooling, and interconnection portfolio to host enterprise generative AI and LLM workloads (private AI and inference zones), supporting customers who deploy GPUs/accelerators and private LLMs; initiative surfaced through CoreSite’s 2024–2025 State of the Data Center materials and product messaging. |
| NVIDIA DGX-Ready Data Center certification and GPU-ready offerings | Participation in NVIDIA’s DGX-Ready Data Center program and explicit certification of specific campuses to host NVIDIA DGX/GPU hardware, enabling customers to deploy large-scale training and inference stacks for generative AI and LLMs. |
| AI inference zones and high-density rack/suite deployments | Creation of AI inference-optimized zones and offers (high-density power, enhanced cooling and direct cloud onramps) designed for low-latency LLM inference and agentic AI workflows, plus multicloud networking (e.g., AWS 400G Direct Connect, Oracle FastConnect 50G) to support distributed AI architectures. |
PE / Ownership History
| Event | Details |
|---|---|
Stonepeak forms joint venture with American Tower (CoreSite DevCo JV) to develop, construct and operate an 18 MW data center in Denver, Colorado; Stonepeak holds an 85% stake, CoreSite/American Tower 15%. July 31, 2024 | Stonepeak invested through its opportunities fund and took a controlling economic stake in the JV to deploy and operate new AI/ML-optimized data center capacity in Denver. |
Stonepeak acquires a material stake in American Tower’s U.S. data center business (CoreSite-related stake purchase reported ~ $2.5B). 2024 | Press coverage reported Stonepeak purchased a sizable economic stake from American Tower to scale data center investments; named PE deal partners were not specified in public announcements for this transaction. |
Acquired by American Tower Corporation for approximately $10.1 billion in cash ($170 per share). November 15, 2021 (announced) / December 28, 2021 (completed) | Transaction was unanimously approved by both boards and completed in December 2021; rationale focused on combining wireless and wireline communications real estate and scaling data center operations. |
Executives
| Name | Background |
|---|---|
Paul E. Szurek President and Chief Executive Officer (CEO) | Long-time CoreSite executive and board member (board since 2010); leads corporate strategy and product/market development for colocation and interconnection services; extensive experience in data center operations and REIT management. |
Hernandez Xu Chief Financial Officer (CFO) | Leads finance, accounting, and financial planning functions for CoreSite; background in corporate finance and financial operations in technology/real estate sectors (public profiles and company announcements list CFO role but provide limited public CV detail). |
Joseph (Joe) Liccardo Senior Vice President, Operations (COO-equivalent) | Oversees customer success, data center operations, network operations and technical programs across CoreSite’s portfolio; ~25+ years of facility and data center operations experience, including leadership roles in large multi-site operations. |

Stonepeak