Why AI Data-Center Delays Are Putting Equinix Stock in Focus, October 2026

Equinix's FY2025 capital expenditure reached $4.31B — up 40.6% from FY2024 — as AI buildout-delay reports coincide with the company's $4B atNorth close and Inference Exchange launch with NVIDIA and Together AI, prompting both an analyst upgrade and a renewed valuation debate.

Reports of delays in the AI data-center buildout reached analytical desks in October 2026, prompting Simply Wall Street to place both Equinix (NASDAQ: EQIX) and Digital Realty under the same scrutinizing lens. The timing is notable. Equinix entered this period having just completed a $4 billion acquisition of Nordic data-center operator atNorth alongside CPP Investments — a transaction whose completion, reported in September 2026, was accompanied by a stock gain — while simultaneously forging a strategic alliance with NVIDIA and Together AI to launch the Inference Exchange, a platform delivering GPU compute access and more than 200 open-source models directly inside Equinix colocation facilities. EQIX shares closed at $987.44, up 1.2%, in the October 20 session; they had held above $1,070 as recently as August 2026 when second-quarter results were published, though no published source has explicitly attributed the intervening move lower to a single cause.

The strategic architecture behind these moves has been in development longer than the recent headlines suggest. The Futurum Group articulated in September 2026 what Equinix has been signaling for some time: a transition from raw-compute colocation hosting toward specialized AI infrastructure. Equinix's CEO reinforced the same thesis in a CNBC interview, arguing that the company's neutral interconnection platform lets enterprises deploy and benchmark competing AI models without being locked into any single provider. That positioning received a market stress-test in September 2026: Traders Union reported the stock fell 1.3% following the distributed AI inference program announcement, while Blockonomi separately reported a 2% gain following what it described as a major NVIDIA partnership announcement in the same month. The two reports may reflect different trading sessions, and no source reconciles them definitively. Separately, Seeking Alpha published an upgraded buy rating in October 2026, citing Equinix's AI capex acceleration and the atNorth consolidation as the basis for increased conviction — a constructive reading of the same set of facts that other analysts are simultaneously scrutinizing for valuation justification.

The financial scale of Equinix's commitment is not in dispute. Slicast's compilation of SEC XBRL data shows capital expenditure reached $4.31B in FY2025, a 40.6% increase from $3.07B in FY2024, representing a capex intensity of 47% against FY2025 revenue of $9.22B. The single largest quarterly outlay in Slicast's 57-period dataset spanning 2010 through 2026 was $1.26B, recorded in the quarter ending March 31, 2026. Within the data-center peer group tracked by Slicast, Equinix ranks first by absolute capital expenditure among three comparable companies; on capex intensity within a four-company comparison, it ranks second, with Digital Realty leading that metric. Second-quarter 2026 revenue came in at $2.6B, and management subsequently raised full-year 2026 revenue guidance; at the Six Five Summit, the company attributed the increase to strong AI-driven demand for high-density colocation capacity. Equinix also raised new debt to fund expansion, per a Simply Wall Street analysis from August 2026. The $1.2 billion agreement to sell Tokyo data centers held in a GIC joint venture to Keppel DC REIT, reported in September 2026, illustrates a parallel discipline of recycling mature assets.

Several secondary developments round out the picture. Equinix and Schneider Electric were reported in September 2026 to be trialing software-defined switchgear for power distribution and cooling management, a technology the two companies characterized as enabling faster deployment cycles and over-the-air upgrades without downtime. On sustainability, Equinix signed an 11.5 MWp solar power purchase agreement with Flo Energy in Singapore, its fourth such deal in Singapore in two years, and secured rights to expand Singapore data centers alongside Digital Realty, STT GDC, and Keppel. The Johannesburg facility was scaled to 24 MW, though a large expansion land plot there reportedly remains undeveloped. Saudi Arabia's central bank was reported in September 2026 to have purchased EQIX shares, marking sovereign-capital entry into the listed colocation sector. Korean AI inference chip maker FuriosaAI deployed inference servers at Equinix's Lisbon data center, targeting Europe's sovereign AI market.

The overall picture is genuinely two-sided. On the opportunity side, the NVIDIA and Together AI alliance, a dedicated AI inference service slated to launch in 2027, and the vendor-neutral architecture that management argues is structurally differentiating all represent tangible upside; CPP Investments' commitment in the atNorth deal and the reported Saudi sovereign equity position suggest institutional confidence in the long-duration infrastructure thesis has not wavered. On the risk side, a 47% capex intensity leaves limited income-statement cushion against any demand shortfall; the new debt layer adds fixed-cost sensitivity; and sector-wide AI buildout delays — the precise concern that triggered October's analytical focus — could push lease signings and inference bookings further into the future than the current valuation multiple assumes. The question Yahoo Finance UK raised in September 2026 — whether AI-driven revenue growth can justify Equinix's earnings premium — remains open, and the split market reaction to the September 2026 inference announcements suggests investor conviction has not fully coalesced. Three signals are worth monitoring closely: the pace of enterprise bookings when the 2027 AI inference service formally opens, and whether the Inference Exchange customer count builds at a rate consistent with the premium multiple; whether South Africa expansion land and other pipeline capacity comes online in line with demand rather than ahead of it; and how quickly the AI-oriented revenue mix grows relative to traditional colocation, since that ratio will ultimately determine whether this capex cycle looks prescient or premature.