Equinix has revised its financial outlook for 2026 and extended its long-term growth targets through 2029, citing robust demand for AI infrastructure, cloud connectivity, and digital colocation services. The company’s second-quarter performance exceeded expectations, reinforcing confidence in sustained enterprise investment despite ongoing capacity constraints in its global data center footprint.
What happened
Equinix reported record interconnection growth and one of its strongest booking quarters to date. The company’s results reflect broad-based demand across sectors, with AI infrastructure playing a central role in driving adoption. While GPU demand often dominates headlines, the underlying trend highlights the critical role of connectivity between enterprises, cloud providers, and digital infrastructure providers in enabling AI workloads.
The company also raised its full-year 2026 guidance and set new long-term targets extending to 2029. These adjustments follow a pattern of consistent growth, supported by expansions in key markets such as São Paulo, where Equinix recently committed $76 million to a new data center. The facility, labeled SP3, will add nearly 2,800 cabinets in its first phase, effectively doubling the company’s capacity in Brazil. Over 800 companies are already colocated across Equinix’s existing Brazilian sites, signaling strong regional demand.
- Equinix raised 2026 financial guidance and set new long-term targets through 2029.
- Record interconnection growth and strong Q2 bookings reported.
- $76M investment in São Paulo data center (SP3) to add ~2,800 cabinets.
- Over 800 companies already colocated in Equinix’s Brazilian facilities.
Why it matters
The results underscore the growing importance of AI-ready infrastructure in enterprise digital strategies. Equinix’s focus on connectivity and colocation aligns with a shift in enterprise AI adoption, where the bottleneck is no longer just model access but the ability to deploy workloads near the right data, with reliable networking and cooling. The company’s expansion of Fabric Geo Zones—network-level controls to keep data within specific geographic jurisdictions—addresses compliance risks for regulated organizations, particularly in hybrid and multicloud environments.
Equinix’s decision to shutter its bare metal cloud service, Equinix Metal, further signals a strategic pivot toward AI-era networking and edge deployments. The move, effective June 30, 2026, ends the company’s short-lived foray into bare metal cloud infrastructure, originally launched following its acquisition of Packet. Remaining customers were required to unwind servers, networking links, and billing records before the shutdown, highlighting the operational challenges of infrastructure transitions.
What to watch
Equinix’s growth trajectory will depend on its ability to navigate regional constraints, such as power and grid capacity limitations in markets like Amsterdam. The company recently secured permission to build a data center in Amsterdam-Zuidoost, though only one of four planned towers will proceed due to local power constraints. The project’s scaled-back scope reflects broader tensions between urban development, housing priorities, and digital infrastructure expansion, with implications for long-term planning through 2036 and beyond.
The company’s collaboration with Cisco, NVIDIA, and Presidio to deploy AI factory designs across its global colocation footprint also merits attention. These partnerships aim to provide enterprises with production-grade AI infrastructure tested in real data center environments, addressing concerns around workload placement, data control, and network performance. As AI adoption matures, Equinix’s ability to deliver integrated, compliant, and scalable solutions will likely shape its competitive positioning in the colocation and interconnection market.
Automated pipeline · Cloud & Infrastructure
Synthesized from 1 industry feed on 31 Jul 2026. Passed independent editor verification (score 85/100) before publication. Style guide v1.4.
Sources
Decision trail
- Checking for duplicates — Deduped batch of 1 candidates
- Checking for duplicates — New story No recent or in-pipeline article covers Equinix's Q2 2026 earnings or AI infrastructure demand.
- Checking for duplicates — New story pre_write:; No recent or in-pipeline article covers Equinix's Q2 2026 earnings or AI infrastructure demand.
- Writing the article — Draft created article_id=381 slug=equinix-lifts-2026-guidance-on-record-ai-infrastructure-demand
-
Editor review — Approved
- Score: 85/100
- Factual grounding: The draft states the bare metal cloud service shutdown was 'effective June 30, 2026,' but Source 1 (Jun 30, 2026) states the shutdown occurs *on* June 30, 2026, implying the date is the event date, not the effective date. The draft should clarify this as the shutdown date, not an 'effective' date.
- Style compliance: The standfirst mentions 'long-term targets through 2029' but does not specify these are *financial* targets, which could mislead readers. The standfirst should clarify the nature of the targets (e.g., 'financial targets').
- No copied phrasing: The phrase 'GPU demand dominates headlines' is nearly identical to Source 1's wording ('GPU demand dominates headlines'). While the idea is common, the phrasing should be restructured to avoid echoing the source.
- Audience relevance and notability: The draft includes a detailed explanation of the Amsterdam data center power constraints, which is relevant but tangential to the core financial/guidance story. While defensible, this section could be tightened to focus more on the financial implications for hosting professionals.
- Generating reader Q&A — Generated 5 items
- Assigning hero image — Reused library image reused image #8
- Linking related stories — Linked 5 relations from 325 candidates
- Publishing — Published equinix-lifts-2026-guidance-on-record-ai-infrastructure-demand
- Mastodon — Posted https://mstdn.social/@hostingpaper/117016807428351600




Discussion · coming soon
Be the first to join the thread when community discussion launches.