Digi Power X (DGXX) Q1 2026 earnings review
Crypto Dies So AI Can Run
Digi Power X is undergoing a radical, expensive transformation. Q1 2026 revenue fell 27% YoY to $6.8M, a deliberate deceleration as the company winds down legacy crypto mining to make room for AI compute. While Adjusted EBITDA reversed to a positive $1.1M, GAAP Net Loss actually widened to $4.7M due to heavy pre-revenue infrastructure spending. The real story is the balance sheet: management amassed $125M in cash with zero long-term debt to fund a massive pivot. With a $1.1B colocation deal signed and NeoCloudz live, the foundation is set. However, projecting $250-$300M in 2027 revenue from a $6.8M current quarter requires near-perfect execution.
🐂 Bull Case
The $1.1B, 10-year colocation agreement guarantees a massive, contracted revenue floor. Coupled with the live NeoCloudz GPU cluster, the pivot from volatile crypto to predictable AI infrastructure is becoming a reality.
$125 million in cash, zero long-term debt, and $15 million in digital assets provide immense runway. The company has already deployed $45M YTD in CapEx, proving they are executing the buildout without relying immediately on debt markets.
🐻 Bear Case
Management's 2027 guidance of $250-$300M means scaling revenue by roughly 10x from current annualized rates. Any supply chain hiccups with NVIDIA hardware or delays in Phase 2 commissioning will crush these targets.
A cash balance doesn't jump from $6M in late 2025 to $125M today without severe shareholder dilution. The lack of long-term debt implies the AI pivot was heavily funded on the backs of equity holders.
⚖️ Verdict: ⚪
Neutral. The strategic pivot is absolutely correct, and the $1.1B contract validates the model. However, the widening GAAP losses, heavy dilution, and astronomical 2027 growth targets leave zero margin for error.
Key Themes
NeoCloudz is Officially Live
The company recognized its first AI revenues in May 2026 via NeoCloudz GPU-as-a-Service. Utilizing an initial fleet of NVIDIA B200 and B300 GPUs deployed at the Columbiana facility, this marks a reversing trend from relying purely on crypto and low-margin energy sales toward high-margin AI compute.
Adjusted EBITDA Inflection
Despite a 27% YoY revenue decline, Adjusted EBITDA reversed from $(1.3)M in Q1 2025 to a positive $1.1M in Q1 2026. This $2.4M swing proves that the wind-down of unprofitable legacy operations is structurally improving the company's core profitability profile.
Columbiana AI Hub Transformation
With $45M in YTD CapEx already deployed mostly at the Columbiana site, Phase 1 commissioning is actively transitioning the company's real estate into a tier-grade AI center. The targeted 90 MW of AI colocation by 2027 is the primary engine behind the new growth narrative.
Unspoken Shareholder Dilution
Management boasts about $125M in cash and zero debt, but this capital didn't materialize from operations. Working capital spiked $68M YoY. Following the $100M At-The-Market program active in late 2025, this cash pile is a direct result of massive share issuance. Investors paid heavily for this balance sheet.
Net Losses Contradict The 'Inflection' Narrative
While management labeled Q1 an 'inflection point' based on Adjusted EBITDA, actual GAAP Net Loss widened heavily, decelerating from $(1.6)M a year ago to $(4.7)M. The positive EBITDA excludes $1.5M in D&A, $1.3M in stock comp, and $3.8M in crypto revaluation losses—indicating the real cost of running the business remains deeply negative.
Corporate Overhead is Surging
The widening net loss was partially driven by 'growth in corporate headcount supporting the AI build program'. As the company opens a Silicon Valley office to recruit top engineering talent, SG&A expenses will likely accelerate sharply before NeoCloudz reaches full utilization.
Other KPIs
Accelerating dramatically. The company deployed $45M into GPU equipment and Phase 1 Columbiana buildout in just the first few months of 2026. For context, they spent roughly $3.0M in the first nine months of 2025. This shows aggressive, real-world execution of the AI pivot.
Accelerating massively. Up from essentially near-zero ($0.8M) a year ago. This provides the necessary war chest to execute the $1.1B colocation agreement and fund the Silicon Valley expansion without relying on high-interest debt.
Accelerating, up 29% YoY. This direct reflection of capitalized investments at the Columbiana facility proves the hardware is arriving and being installed.
Guidance
Accelerating exponentially. This represents a monumental leap from the $6.8M generated in Q1 2026. The guidance assumes seamless execution of 90 MW in colocation and massive GPU-aaS uptake.
Accelerating. The base $80-$100M is de-risked by the new 10-year Colocation Agreement covering Phase 1 and partial Phase 2 operations. The remaining $100M upside requires leasing out an additional 50 MW.
Accelerating. Expected to scale to ~10 MW over the course of the year. This segment carries the highest margin potential but also the highest utilization risk if AI compute demand softens.
Key Questions
ATM Issuance Reality
To achieve the $125M cash balance without taking on long-term debt, how many total shares were issued under the At-The-Market program, and what is the current fully diluted share count?
Phase 2 Capital Requirements
With $45M already spent YTD on Phase 1, what is the anticipated capital expenditure required to fully commission the targeted 40 MW Phase 2 rollout by 2027?
Colocation Margin Profile
On the newly signed $1.1B, 10-year colocation agreement, what is the expected gross margin profile, and how much of the underlying power costs are passed directly to the client?