WhiteFiber (WYFI) Q1 2026 earnings review

Infrastructure Boom Masks Exploding Corporate Overhead

WhiteFiber is successfully positioning itself as a vital landlord for the AI boom, but the cost of scaling is devastating its bottom line. Revenue grew 31% YoY to $21.9 million, driven by a 190% explosion in Colocation services as the MTL-3 facility came online. However, management's celebration of 'positive adjusted EBITDA' ignores a massive profitability reversal: Net Income swung from a $1.4 million profit last year to a $12.0 million loss. General and Administrative (G&A) expenses quadrupled, eating up 81% of total revenue. The bull case rests entirely on an enormous $921 million in Remaining Performance Obligations (RPO) waiting to be unlocked.

๐Ÿ‚ Bull Case

Massive Contract Backlog

WhiteFiber is sitting on $921 million in Remaining Performance Obligations (RPO) for colocation services. This provides multi-year revenue visibility entirely insulated from short-term GPU cloud pricing wars.

Power Bottlenecks Solved

Securing utility power is the hardest part of AI infrastructure. Duke Energy has completed the delivery of 54 gross megawatts to the NC-1 campus, de-risking the upcoming 40-megawatt Nscale deployment.

๐Ÿป Bear Case

Uncontrolled Overhead Costs

G&A expenses surged 318% YoY to $17.8 million, driven by public company costs and $7.3 million in share-based compensation. Revenue growth cannot offset overhead scaling this aggressively.

Execution and Supply Chain Risks

A newly identified supply-chain issue with medium-voltage switchgear components at NC-1 threatens to delay the timeline. If the Q3 2026 target for full revenue contribution slips, cash burn will intensify.

โš–๏ธ Verdict: โšช

Neutral. The underlying assets and $921 million RPO validate the business model, but corporate bloat and soaring interest expenses are destroying near-term equity value. Execution at the NC-1 facility is the sole catalyst that matters now.

Key Themes

DRIVERNEW๐ŸŸข

Colocation Eclipsing Cloud Services

The business model is undergoing a massive shift. While Cloud services revenue grew a modest 13% YoY, Colocation surged 190% to $4.8 million following the October 2025 operational launch of MTL-3. With $921 million in colocation RPO on the books, WhiteFiber is transitioning from a GPU cloud vendor into a pure-play AI infrastructure landlord.

CONCERNNEW๐Ÿ”ด๐Ÿ”ด

G&A Expense Spiral Contradicts Profitability Narrative

Management touted 'positive adjusted EBITDA,' but this metric conveniently excludes a massive spike in operating costs. General & Administrative expenses skyrocketed from $4.2 million in 25Q1 to $17.8 million in 26Q1. Stripping out $7.3 million in share-based compensation still leaves an aggressive cash-burn profile that wiped out the company's 60% gross margins.

DRIVER๐ŸŸข

Securing the AI Power Bottleneck

The macro AI narrative has shifted from 'securing GPUs' to 'securing power.' WhiteFiber successfully had Duke Energy deliver an initial 54 gross MW of utility power to the Madison, NC site. This physical power availability is the ultimate moat in the current high-density AI infrastructure market.

DRIVERNEW๐ŸŸข

Cloud Pivot to Capital-Light Structures

WhiteFiber is altering its cloud strategy to reduce CapEx risk. The new $17 million, two-year agreement with Hyperbolic (supporting Modal Labs) exclusively utilizes WhiteFiber's existing fleet of H200 GPUs. Repurposing existing hardware without requiring incremental GPU capital expenditures is a critical step toward improving the cloud segment's cash yield.

CONCERNNEW๐Ÿ”ด

Supply Chain Snags at Critical NC-1 Campus

Management flagged a 'recently identified supply-chain-related issue affecting certain medium-voltage switchgear components' at the NC-1 site. While they still expect capacity delivery to begin in Q2 2026, any delay to the Q3 full-revenue timeline directly threatens the company's ability to service its growing debt load.

CONCERNNEW๐Ÿ”ด

Surging Interest Expenses and Debt Burden

Interest expenses hit $2.0 million in Q1, up from zero a year ago. WhiteFiber recently raised $230 million in convertible notes and drew another $18 million from an Icelandic credit facility. Until the NC-1 site starts generating cash, the cost of carrying this debt will severely suppress net income.

Other KPIs

Gross Margin60.2%

Stable. Gross margin excluding D&A remained effectively flat compared to 60.5% in 25Q1. This proves the core unit economics of both the cloud and colocation businesses are healthy; the profitability issues are entirely driven by below-the-line corporate overhead and financing costs.

Adjusted EBITDA$3.0 million

Decelerating violently. Down 50% from $6.0 million in 25Q1. Despite a 31% increase in total revenue, the weight of operating expenses slashed core operating cash profitability in half.

Cash and Cash Equivalents$75.8 million

Bolstered significantly by the $230 million convertible note issuance in Q1. The company is actively front-loading cash to guarantee the completion of the NC-1 campus and the acquisition of the MTL-3 site (which closed in May).

Guidance

NC-1 Revenue ContributionInitial Q2 2026 / Full Q3 2026

Accelerating. The 40 MW IT load deployment with Nscale is the company's most important catalyst. If full revenue begins in Q3 2026 as guided, the total company revenue profile will step-function upward, unlocking the $921 million RPO backlog.

Hyperbolic Cloud Contract$17 million TCV

Accelerating. Expected to begin contributing to cloud revenue in June 2026. Because this leverages existing H200 GPUs, it should drop highly accretive margins to the bottom line in the second half of 2026.

Key Questions

Normalized G&A Run Rate

G&A expenses were $17.8 million this quarter, up 318% YoY. How much of this is one-time public company structuring versus the permanent run-rate required to manage your expanding data center footprint?

Switchgear Delay Impact

Can you quantify the exact timeline risk posed by the medium-voltage switchgear supply chain issue at NC-1? If components are delayed past Q2, what are the penalty structures with Nscale?

Cloud Segment CapEx

The Hyperbolic deal utilizes existing H200 GPUs. Should we assume GPU capital expenditures will remain minimal going forward, shifting entirely toward infrastructure build-outs?