This article first appeared on GuruFocus .
-
Consolidated Revenue:$1 million in Q2 2026, compared to $0.5 million in Q2 2025 and $1.4 million in Q1 2026.
-
Accelsius Revenue:Contributed $0.9 million, or 96% of total consolidated revenue.
-
Net Loss:$34.9 million for the quarter, compared to $27.8 million in Q1.
-
Adjusted EBITDA:Loss of $22.6 million, versus an $18.4 million loss in Q1.
-
General and Administrative Expenses:$14.5 million, down 22% compared to Q2 2025.
-
Cash and Restricted Cash:$46.5 million at quarter end, compared to $60.4 million at the end of Q1 (which included $5 million of restricted cash).
-
Cash Flow:Year-to-date, used $59.5 million in operating activities and generated $41.6 million from financing activities.
-
Capital Raised:Approximately $13 million through draws on standby equity purchase agreement at an average price of $6.21 per share.
Release Date: August 13, 2026
For the complete transcript of the earnings call, please refer to the full earnings call transcript .
Positive Points
-
Innventure Inc ( NASDAQ:INV ) reported a 100% year-over-year increase in consolidated revenue for Q2 2026, reaching $1 million.
-
AeroFlexx's commercial pipeline grew 9% quarter-over-quarter to nearly $35 million, with new partnerships in Latin America and Europe.
-
Accelsius published a major third-party validation showing its NeuCool system runs GPUs 9-14C cooler than single-phase, using a third of the coolant flow.
-
Accelsius is now focusing on hyperscalers, chip makers, and server OEMs, with active proof-of-concepts and deep evaluation cycles with several key industry leaders.
-
The company raised approximately $13 million through its standby equity purchase agreement at an average price of $6.21, strengthening its balance sheet.
-
Innventure Inc ( NASDAQ:INV ) reduced general and administrative expenses by 22% year-over-year in Q2 2026.
-
The company is pursuing a strategy to finance AeroFlexx and Refinity at the operating company level, minimizing dilution for Innventure shareholders.
Negative Points
-
Innventure Inc ( NASDAQ:INV ) suspended its revenue targets for Accelsius due to market constraints, including power availability, GPU access, and site allocations for smaller early adopters.
-
The DarkNX purchase order was removed from the 2026 forecast after the development site became unavailable, impacting near-term revenue expectations.
-
Accelsius is no longer expected to reach cash flow breakeven this year, and Innventure Inc ( NASDAQ:INV ) has withdrawn its 2028 consolidated positive cash flow target.
-
The company's net loss widened to $34.9 million in Q2 2026, up from $27.8 million in Q1, with adjusted EBITDA loss increasing to $22.6 million.
-
Cash and restricted cash decreased to $46.5 million at the end of Q2, down from $60.4 million at the end of Q1, reflecting significant operating cash use.
-
Innventure Inc ( NASDAQ:INV ) anticipates a need for additional capital at the parent level due to the extended timeline for Accelsius to achieve positive cash flow.
-
The company acknowledged that bookings and revenues will remain lumpy and hard to predict until two-phase adoption is established, with no material updates expected every quarter.
Q & A Highlights
Q: Can you provide more context on the DarkNX purchase order setback? Was it related to power availability, financing, GPU availability, or something else? A: John Hewitt, CEO of Accelsius, explained that DarkNX lost the original development site due to issues with the power envelope. He confirmed that the order is expected to transfer to a new site, but it has been removed from the 2026 forecast because finding an alternate location and securing allocations will take time, making the timing uncertain.
Q: How advanced are the discussions with hyperscalers regarding an executed statement of work (SoW), and what does the timeline look like from securing one to commercial deployment and revenue? A: John Hewitt stated that there are multiple hyperscale conversations in various stages, with very active discussions happening. He outlined that the typical cycle from an early SoW to volume deployment runs between two and three years, involving technology validation, proof-of-concept deployments (sometimes as large as a row in an existing data center), and then integration into the data center reference design.
Q: Given the revised outlook, is it too early for any visibility into 2027 as a meaningful commercialization year for Accelsius? A: John Hewitt confirmed that it is too soon for specific revenue and bookings visibility into 2027. He emphasized that the company is heavily focused on achieving the four key milestones outlined in the call: chip maker engagement, server OEM/ODM co-development, moving beyond proof-of-concept with a hyperscaler, and delivering benchmark data with leading thermal labs.
Q: Can you expand on the milestones for getting integrated with a server OEM? Can a hyperscaler make that happen faster, and are they the real pivot point? A: John Hewitt explained that Accelsius is working with server OEMs and ODMs in parallel. While these partners may do engineering work in advance, they move much quicker when there is a demand signal from a hyperscaler or Neocloud. He confirmed that end-customer demand will significantly motivate OEMs and ODMs, so the switch could flip faster, but the timing remains uncertain.
Q: Can you talk about the potential to raise capital at the company level and the puts and takes regarding cash runway and burn? A: CFO Dave Yablunosky stated that the company has $41.5 million in cash on the balance sheet and access to multiple avenues to raise cash, including the standby equity purchase agreement. He acknowledged that with the revised timeline for Accelsius, there could be a need for additional cash raises in the second half of the year, but emphasized the company will be opportunistic to minimize dilution and preserve shareholder exposure to Accelsius.
Q: Are there any updates on the ability for AeroFlexx and Refinity to do something strategic or become self-funding? A: CEO Bill Haskell noted that both Refinity and AeroFlexx have initiatives underway to raise their own capital and become self-funding, which minimizes the need for parent-level cash. He added that AeroFlexx is "turning the corner" with a growing pipeline and larger CPG partners, though these companies move slowly. He suggested there are various avenues available for AeroFlexx depending on how the rest of the year goes.
Q: When do you expect the NVIDIA Rubin (Fenman) architecture and its cooling architecture to be announced? A: John Hewitt admitted he does not know the exact timing of the announcement, but noted the company is monitoring it very carefully. He acknowledged that the design gets locked down well ahead of the announcement, but could not provide more specific details.
Q: When did you guys go from "if" to "when" regarding two-phase adoption, and what was the catalyst? A: John Hewitt explained that internally, the conviction that it's "when, not if" came relatively early when the company proved the technology worked and got strong proof points on performance versus single-phase and air cooling. The shift to broader market conviction is driven by the rest of the ecosystem recognizing the same. Bill Haskell added that industry projections for the two-phase market in 2030 have grown to $9 billion, materially higher than a year ago, and that larger players are moving in sooner, bringing the inflection point closer.
For the complete transcript of the earnings call, please refer to the full earnings call transcript .
