INNVENTURE, INC.
| Revenue | $2.4M |
|---|---|
| Operating income | $-58.8M |
| Net income | $-47.4M |
| Free cash flow | $-60.5M |
| Operating margin | -2453.5% |
| Net margin | -1977.5% |
| Return on equity | -22.3% |
| Period | 2026 |
OpenFilings analyst
Our analyst
Agentic read of the latest earnings call — recommendation updates when we process a new transcript. Research synthesis, not investment advice.
avoid buying for now; this is a high-burn, dilution-dependent commercialization option rather than a de-risked growth story.
Q2 revenue fell sequentially to $1.0 million from $1.4 million, adjusted EBITDA loss widened to $22.6 million, cash and restricted cash declined to $46.5 million, and management suspended Accelsius revenue/cash-flow targets while withdrawing its 2028 consolidated positive-cash-flow target. The long-term NeuCool case has credible technical validation and a potentially large market, but there is still no disclosed hyperscaler agreement, material deployment, or visibility into 2027 revenue.
Near term
Accelsius has removed the DarkNX project from its 2026 forecast after the customer lost its site because of power-envelope constraints. Management expects a possible transfer to another site, but timing and contract certainty are unclear.
Near-term results are likely to remain volatile: management explicitly warned that bookings and revenue will be lumpy and difficult to predict as smaller early adopters lack power, GPU allocations, and site capacity.
Cash preservation and financing are immediate stock drivers. The company used $59.5 million in operating cash year to date, raised approximately $13 million through its standby equity purchase agreement at an average $6.21 per share, and said additional parent-level capital may be needed in the second half.
The October 1 CEO transition adds execution risk, although the successor is an internal executive with prior responsibility for Refinity and Accelsius board involvement.
AeroFlexx's $35 million pipeline and Refinity's planned 10-kiloton engineering design are positives, but neither currently offsets Accelsius' cash needs or has been translated into dependable consolidated revenue.
Longer term
NeuCool's third-party test showed GPUs running 9–14 degrees Celsius cooler than the factory single-phase system, with roughly one-third the coolant flow, and management cited an average 5% GPU-capacity advantage within the same power envelope. These are meaningful technical proof points, but they remain a limited validation rather than evidence of scaled commercial adoption.
The revised Accelsius strategy is aimed at chip manufacturers, server OEMs, server ODMs, and hyperscalers. The key value inflection would be reference-design inclusion, co-development and warranty integration, or an executed hyperscaler statement of work—not additional small-customer bookings.
The opportunity could expand as chip power rises and data centers face power and water constraints; management cited a projected $9 billion two-phase cooling market within a $30 billion liquid-cooling market by 2030. The investment case depends on Accelsius securing ecosystem position before larger cooling vendors do.
The moat remains unproven. Large customers are evaluating multiple cooling architectures, and the company must convert technical performance into reliability, factory integration, server warranty coverage, and procurement approval.
Hyperscaler commercialization can take two to three years from early statement of work through proof of concept, reference design, and volume deployment, making the path to meaningful revenue substantially longer than the previously implied 2027 commercialization thesis.
Red flags
Management replaced revenue guidance with milestone reporting after missing the prior expectation for Accelsius to approach a roughly $100 million annualized revenue run rate and cash-flow breakeven by year-end. This reduces accountability and leaves valuation anchored to aspiration rather than financial visibility.
Analysts could not obtain the portion of the approximately $50 million of cited Accelsius bookings attributable to DarkNX, nor confirmation that the order will automatically transfer. Management only said it expects the order to transfer after a new site and allocations are secured.
Management gave no visibility into 2027 revenue and could not provide a timetable for the Fenman cooling architecture announcement, despite its potential importance to customer design decisions.
The independent integrator behind the validation was not named, the customer base was described only generally, and the test involved one Dell/NVIDIA server configuration. Broader reliability, maintenance, warranty, and total-cost-of-ownership evidence is still needed.
Liquidity commentary was not fully consistent: the prepared remarks cited $46.5 million of cash and restricted cash, while the Q&A referenced $41.5 million of cash. Against $59.5 million of year-to-date operating cash use, the company acknowledged a likely need for further capital and dilution.
The bullish claim that the market has shifted from 'if' to 'when' was not matched by disclosed purchase commitments from hyperscalers, chip makers, or server manufacturers. Technical superiority alone may not overcome entrenched single-phase suppliers and long qualification cycles.
OpenFilings analyst view from primary-source filings and earnings calls — not investment advice.
Upcoming earnings
Earnings transcripts
Press & signals
- GlobeNewswire
Innventure Announces Establishment of At-the-Market Equity Program and Discontinuation of SEPA
- GlobeNewswire
Innventure Leadership Increases Focus on Accelsius Value Creation, Implements Significant Cost Cuts, and Strengthens Board Independence
- GlobeNewswire
Innventure Announces Board Leadership Changes to Further Enhance Board Independence and Acceleration of CEO Transition
The rest of this company's wire — with summaries and source links — is part of Pro.
See plans