Innventure - Earnings Call - Q1 2025

Executive Summary

• Q1 2025 revenue was $0.224M, below S&P Global consensus of $0.58M as management reiterated that revenue growth is expected to inflect in 2H 2025. Consensus from S&P Global shown for context; see Estimates Context section for details.*
• On S&P’s “Primary EPS,” INV printed 1.77 vs -0.19 consensus, but GAAP diluted EPS was -$3.10, driven by a non‑cash $233.2M goodwill impairment; adjusted EBITDA was -$21.8M. Consensus from S&P Global shown for context; see Estimates Context section for details.*
• Operating company momentum was the focus: management highlighted accelerating engagement for Excelsior (two‑phase liquid cooling), a ~300% spike in leads since late Feb, a ~200% expansion in the partner network YTD, and proposals now averaging $2–$4M—setting the stage for 2H revenue ramp.
• Liquidity actions post‑quarter added $27M gross proceeds via two convertible debentures ($18M on Apr 14, $9M on May 15), and founders converted ~$18M of related‑party debt to preferred equity, reducing annual interest expense by ~$3M.

What Went Well and What Went Wrong

• What Went Well
  • Management reiterated “confidence in achieving revenue growth inflection during the second half of 2025” and highlighted strong strategic positioning of Accelsius/Excelsior in two‑phase direct‑to‑chip liquid cooling.
  • Commercial traction signals: lead generation up ~300% since late Feb; strategic partner network up ~200% YTD; average proposal size moved from single‑unit POCs to $2–$4M, indicating pipeline maturation.
  • Manufacturing readiness: internal scale seen sufficient to reach profitability at ~100 racks/month, with a global contract manufacturer able to handle 10,000+ racks today; typical large orders could be fulfilled within ~90 days, supporting a 2H ramp narrative.
• What Went Wrong
  • Q1 revenue of $0.224M (largely management fees) missed S&P consensus ($0.58M) and declined sequentially vs Q4 as revenue remains back‑half weighted. Consensus from S&P Global shown for context; see Estimates Context section for details.*
  • GAAP loss per share of -$3.10 was impacted by a $233.2M non‑cash goodwill impairment tied to share price/market cap declines during late Feb–Mar; without the impairment, adjusted EBITDA still reflected a -$21.8M loss on ramp costs.
  • Q&A and commentary acknowledged a slower industry first half (tariff uncertainty and re‑orientation around Nvidia’s roadmap), pushing orders/evaluations later in 2025; management emphasized momentum into 2H.

Transcript

Operator (participant)

Hello, and welcome to Innventure First Quarter 2025 Earnings Conference Call. At this time, all participants are in a listen-only mode. After the speaker's presentation, there will be a question-and-answer session. To ask a question during the session, you will need to press star one one on your telephone. You will then hear an automated message advising your hand is raised. To withdraw your question, please press star one one again. I would now like to turn the conference over to Lucas Harper. Sir, you may begin.

Lucas Harper (CIO)

Thank you, Operator, and thank you all for joining us for Innventure's First Quarter 2025 Earnings Call. My name is Lucas Harper, Innventure's Chief Investment Officer, and joining me from the company are Bill Haskell, Chief Executive Officer, and Dave Yablunosky, Chief Financial Officer. Earlier today, we issued a press release announcing our financial results, which is available on our Investor Relations website along with a supplemental slide presentation. As referenced on Slide 5, we will be discussing non-GAAP financial measures during this call. The most directly comparable GAAP financial measures and a reconciliation of the differences between the GAAP and non-GAAP financial measures are available in our earnings release and supplemental slide presentation on our website. In addition, certain statements being made today are forward-looking statements that are based on management's current assumptions, beliefs, and expectations concerning future events impacting the company.

These forward-looking statements involve a number of uncertainties and risks, including but not limited to those described in our earnings release, Form 10-Q for the period ending March 31, 2025, and other filings with the SEC. The actual results of operations and financial condition of the company could differ materially from those expressed or implied in our forward-looking statements. I would like to hand the call over to Bill.

Bill Haskell (CEO)

Thanks, Lucas, and thanks to everyone joining us today. Given the recency of our last update, we'd like to use today's call to shift our focus from specific first-quarter developments to a high-level recap of the exciting opportunities that lie ahead for Innventure, with a spotlight on Accelsius. AeroFlexx and Refinity continue to execute on their respective strategies, and we are excited to share more about both as we move through 2025. Given the number of inbounds we field from investors on Accelsius, we felt today's call would be a good opportunity to speak at more length about the company. At the highest level, Accelsius has tremendous momentum, but it is important to provide additional context on the compelling market opportunity that underpins a large part of Innventure's value creation potential.

Today, we're excited to have Dino Foderaro, Accelsius' Chief Revenue Officer, on the call to provide an insider's perspective on Accelsius' market opportunity, technology differentiators, and adoption momentum. With nearly 10 years in the data center industry, first in power distribution and quality, and now over the last 2.5 years in advanced cooling with Accelsius, Dino has had a front-row seat to the infrastructure challenges shaping today's market. As CRO, he leads the client operations group, applying his operations background to driving execution and ensuring a premier customer experience across every engagement. As we've repeated often, Innventure's goal is to build companies that we believe can achieve a minimum of $1 billion in enterprise value. We'd like everyone to walk away from today's call with a higher level of context and understanding about why Innventure believes in Accelsius' visibility to meet and exceed that milestone.

Hearing directly from Dino, who speaks to our partners and potential customers every day, seems like a good way to provide that color. With that, I'd like to turn the call over to Dino.

Dino Foderaro (CRO)

Thanks, Bill. Thanks to the entire Inventure team for having me on the call. There are many exciting things happening at Accelsius, and today, I'll focus my remarks on three key areas. One, Accelsius' opportunity within the data center liquid cooling market. Two, the technology differentiators of our NeuCool two-phase directed chip technology. Three, a deep dive into the momentum we are developing with ecosystem partners and high-value end users. Let's start with the sizable, rapidly growing market that Accelsius' technology and product set address. This highly attractive market was one of the key factors in Inventure's decision to commercialize the two-phase directed chip technology and one of the many reasons I joined the company. There are a few key secular tailwinds happening in the data center industry that create a compelling backdrop that Accelsius aims to capitalize on.

First, it is no secret that technology companies, and more importantly, companies across all industries, have publicly talked about significantly increasing their data center budgets over the next several years to keep up with the advent of AI and rapidly evolving chip technology. According to the Del Oro Group, worldwide data center spending was roughly $450 billion in 2024, and that number is expected to grow to over $1 trillion by 2029. This creates an incredible secular tailwind for the entire data center ecosystem and especially for critical infrastructure providers such as Accelsius. Second, the overwhelming majority of the nearly 12,000 data centers spanning the globe are still utilizing highly inefficient and antiquated air-cooled solutions, and single-phase water cooling technologies represent a single-digit percentage of cooling installations.

In the short term, single-phase water-based technologies are expected to expand their share of the liquid cooling market as data centers install infrastructure for this technology. However, these solutions have significant limitations. We believe that Accelsius' NeuCool technology is superior to single-phase water-based technologies. For data center operators already using single-phase, switching to Accelsius' NeuCool products does not require a seismic shift or costly rip-and-replace approach. Companies that currently use single-phase cooling solutions can easily de-risk and future-proof their data centers by swapping out their single-phase solutions for Accelsius' NeuCool technology. This allows them to adopt a more efficient and safer solution while utilizing existing infrastructure. Efficiency is among the future-proofing benefits of our solution.

Data center operators and end users that currently employ traditional air cooling can experience a drastic improvement in power usage effectiveness, or PUE, typically reducing the amount of power used for cooling by more than 80%. While the transition from air to liquid cooling can be a bit of a journey, the simpler transition from single-phase water-based technologies to the NeuCool solution unlocks additional savings in the form of reduced HVAC load and increased free cooling. In fact, test data captured in our lab paired with climate data provided by ASHRAE show that customers utilizing our two-phase technology in comparison to single-phase technology in hard-to-cool locations such as Southeast Asia and the Middle East can see tremendous benefits.

Specifically, this study indicated that our NeuCool technology allows AI data center customers in Singapore the ability to utilize free cooling 97% of the time in comparison to 34% of the time with single-phase technologies. Now, let's discuss how data center operators are thinking about the future. It is widely understood across the data center ecosystem that air cooling solutions are largely insufficient for existing and future generations of processors. Based on feedback we received from key customers and early single-phase deployments, we believe that data center operators are increasingly recognizing the benefits of leapfrogging single-phase cooling altogether as rack densities increase. Turning directly to two-phase directed chip solutions like our NeuCool platform can help data center operators meet sustainability targets, uphold service-level agreements, or SLAs, and protect ROI.

The newest AI-focused chipsets expected to enter the market in late 2025 and 2026, such as NVIDIA's Rubin and AMD's MI350, will require liquid cooling. We believe this dynamic will drive accelerated adoption. To demonstrate the drastic rack density impact of these ultra-high thermal design power processors, a quick look into NVIDIA's five-year product roadmap shows near-term rack densities anticipated to eclipse 250 kW per rack and rapidly approach 600 kW, while other chip and server OEMs are targeting configurations landing at 1 MW per rack before 2030. The so-called thermal wall is very real. According to guidelines put forth by ASHRAE, the industry has already extended beyond the abilities of air cooling and is approaching the limitations of single-phase water-based directed chip cooling technologies at breakneck speed.

Today, the current liquid cooling market is estimated at $1.3 billion, and it is growing at an annual rate of approximately 30% according to the Del Oro Group, driven by advancements in AI and chip technology, as I just mentioned. The Del Oro Group estimates that the market will grow to a $5 billion market by 2028. This represents a significant unmet market need, and Accelsius, armed with our NeuCool technology, has positioned itself at the forefront of the future adoption cycle. Studies show our NeuCool solution is superior to current single-phase directed chip systems and benefits a data center operator in five key ways. The first, superior heat removal. Our technology offers 6-8 degrees Celsius more headroom, which unlocks a myriad of benefits for data center operators and end users.

These benefits include increased potential for heat reuse applications, reduced CapEx associated with costly infrastructure, and the ability to handle future generations of processors with increased confidence. Two, use of non-conductive dielectric fluid. Unlike leaks from water-based cooling systems that can cause millions of dollars in damage to IT equipment, studies show that if a leak of our dielectric refrigerant were to occur, it would not result in IT equipment damage. To put the impact of leaks in context, current-generation GPUs can cost upwards of $60,000 each and carry four- to five-month lead times, which can impact an operator's or end user's realized return on investment. Number three, warranty. We provide peace of mind with two- and five-year limited warranty options and offer additional coverage that replaces server electronics if damaged due to leaks of our refrigerant.

Additionally, our relationships with server OEMs and integrators allow us to warranty Accelsius-enabled IT equipment, eliminating a large hurdle to adoption of our NeuCool technology. Number four, lower OpEx. Our technology can run in an environment with up to 45 degrees Celsius facility water supply dependent on processor and rack configuration compared to 27-32 degrees Celsius facility water for single-phase technologies. This allows operators to run chillers less and utilize free cooling, which can provide data center operators up to 4% energy savings per 1 degree Celsius increase, according to data published by Vertiv. Number five, ease of maintenance. All of our systems are built with hot-swappable components that do not require a licensed technician to manage. The systems also do not need to be constantly monitored or flushed, which eliminates costly maintenance services, reduces on-site staff, and minimizes system downtime.

Today, Accelsius has three products, with several more in development. Rather than just technology, we offer a full product set, which provides flexibility for data center environments regardless of use case, location, or scale. Our flexible solutions with in-rack and multi-rack form factors can be deployed within existing data center infrastructures, brownfield sites, without significant cost or downtime, and give data center design firms a full portfolio of highly efficient solutions to draw from when designing new facilities or greenfield sites. We have included specific product details within the slides accompanying today's presentation. The most important thing to highlight between the various solutions is that each is designed to address a specific market participant need, and each product represents a platform that can and will evolve as the market continues to mature.

From a competitive perspective, we believe there is only one other player in the market today that offers a viable two-phase directed chip liquid cooling technology, and Accelsius is in a lead position. In a multiple billion-dollar market such as this, even capturing a very small percentage of the market share would lead to tremendous growth and success. Now, I'd like to frame the magnitude of the two-phase liquid cooling adoption cycle and how that could translate to results. As you've heard from the Inventure team, Accelsius is focused on building relationships with four key groups: hyperscalers and multinational OEMs, global resellers, colocation providers, and AI-as-a-service provider, also referred to as Neoclouds. Today, I will focus on the available opportunities with hyperscalers, multinational OEMs, and our global resellers. First, hyperscalers regularly require large quantities of racks, with some needing approximately 1,000 racks per week.

To provide an illustration for the true size of the opportunity, a single 1,000-rack order would translate into revenues approaching nine figures. That addressable market presents a truly compelling opportunity available to us. Not only is the opportunity real, the engagement we have with the hyperscalers is real. We are currently deep in discussions with most of the major players, having met some of the companies well over 10x. They understand the value the technology provides and know that broader adoption is only a matter of time. Winning the confidence and volume orders from a hyperscale customer is a deliberate multi-phase process that requires alignment across numerous internal stakeholders, with each evaluating how our technology supports their technical architecture roadmap. These are not everyday decisions. They're made in defined windows tied to platform refresh cycles and require detailed reviews, technical validation, and multi-party sign-off.

The process typically begins with lab testing, followed by formal proof-of-concept deployments that run 60-120 days and may involve several iterations. From there, a pilot deployment is initiated to evaluate performance under real workloads and validate manufacturing and supply chain readiness. The questions we're fielding now focus on performance data, POC site selection, test plans, and execution timelines. Clear signs we've moved from interest to evaluation. While orders take time in this space, we are seeing steady traction, strong engagement, and continued momentum as our technology proves itself in the field. Second, let's review our focus on the global OEMs that design, build, and supply the infrastructure technology used to support data center operations. OEMs represent attractive partners given their scale, reach, and extensive customer networks.

It should be noted that these global OEMs generate billions of dollars of revenue and require significant volumes to have a meaningful impact on their respective businesses. Accelsius has a current OEM relationship to white-label a solution, which includes modest minimum volume commitments. That said, our OEM partner, like other global players, typically does not white-label a product for subscale volumes. They white-label solutions that can scale with market demand, and we believe Accelsius is well-positioned to fulfill that need when it arises. What is going to drive that market demand? As I touched on earlier, NVIDIA recently rolled out their GPU roadmap for the next several years, which shows extreme anticipated increases in rack densities. The necessary computing power needed to enable AI is undeniable. Constituents across the data center ecosystem need to act quickly and need to act boldly to stay ahead of this wave.

Accelsius is doing just that. Accelsius' current offering of solutions has been purpose-built to handle the increasing rack densities anticipated well into 2026. We also have a sophisticated roadmap under development that aims to handle projected densities out to 2027 and beyond. Our deep technical team is focused on tackling the most important challenges facing data center operators and end users, which is what drives our impressive roadmap. The team is working on improved vaporator designs that enable extreme heat fluxes and challenging processor designs, evaluating developments in refrigerants and new working fluids, and developing joint solutions with HVAC equipment manufacturers to allow convenient adoption of the NeuCool family of products. NVIDIA's rollout has coincided with an explosion of interest in Accelsius over the last 60 days or so.

Marketing lead generation, where leads are defined as touches or interactions with market participants, has spiked over 300% in 2025 compared to the prior trailing four-month period. These leads are coming in from a diverse set of data center constituents spanning the four main groups we are focused on: hyperscalers, OEMs, resellers, colocation operators, and AI-as-a-service operators. We've also experienced a notable uptick in our strategic partner network, which has grown by close to 200% since the start of 2025. In 2025 alone, we've added partners such as Wesco, Global Switch, Telehouse, Park Place Technologies, and others are engaged in nearing agreement. We expect several meaningful announcements in the coming weeks and months. Now, over half of the opportunities that come to us originate from a selling partner or manufacturing rep, which demonstrates the value of our strategy.

We believe that focusing on the four key groups I outlined above gives us the reach and scale needed to drive significant growth. For example, partnerships with value-added distributors such as Avnet and Climb provide access to approximately 4,000 VARs globally, which simply cannot be replicated using a rifle shot approach and increases the velocity with which we can service new clients. In addition to the ramp and the number of inquiries we have received, the proposals we are generating are increasing both in total scope and average price. The average proposal size at the end of 2024 was typically for single-unit proof-of-concept systems. Today, proposals average between $2 million and $4 million, which demonstrates Accelsius' inclusion in full-scale production opportunities. Let's now discuss Accelsius' ability to meet these levels of demand.

One question we've frequently heard from investors is, "What's Accelsius' current manufacturing capacity, and how does that translate into revenue and profit?" We anticipate having sufficient internal manufacturing capacity to reach profitability, which we believe would happen by delivering around 100 racks per month. Clearly, the volume potential we've outlined here based on the size of the addressable market far exceeds that 100-rack-per-month threshold. While we aren't providing detail on longer-term unit economic expectations, we do expect to achieve significant operating leverage if we receive orders that ramp into the thousands and tens of thousands of racks. This should give investors a good sense of why Innventure has such conviction in the economics and long-term growth opportunity for the business.

To position the company to meet anticipated market needs, Accelsius is focused on partnering with large, global contract manufacturers that can support the much larger orders we would expect in the future. We currently have one such relationship today, and although we can't disclose specifics like the company name or anticipated volume, we can confirm it is with a well-known contract manufacturing company whose primary focus is the tech hardware industry and who would be able to handle a 10,000+ rack order if received today. When Accelsius receives a large-scale order, fulfillment planning is already well underway. Our supply chain team operates with a proactive readiness model, aligning internal and contract manufacturing capabilities to meet volume and timeline expectations. With a North American-focused supply chain and scalable production capacity, both internal and external, we are positioned to deliver at pace.

In most cases, orders of this nature could be fulfilled within a 90-day window, with revenue recognized upon delivery. This timeline adds to our conviction that the conversations Accelsius is having today could lead to strong revenue growth in the back half of 2025. Thanks again for having me on today's call. We at Accelsius are incredibly proud of the hard work our team puts forth each and every day, and we are excited about the future. With that, I'd like to turn the call back over to Bill.

Bill Haskell (CEO)

Thank you, Dino. This is a truly exciting opportunity. Before passing it to Dave, let's recap the key takeaways we hope investors will walk away with. Number one, the liquid cooling market is sizable and growing and is estimated to reach $5 billion within the next three years.

Two, given the trajectory of increasing chip densities, industry players believe adoption of liquid cooling will be all but required, and Accelsius is at the forefront with the leading technology in the space. Number three, the company is well-positioned through their contract manufacturing partnerships to capture the wave of liquid cooling adoption and fulfill potentially significant demand. Four, that demand from the large global players we are engaged with far exceeds what Accelsius needs to become a profitable and rapidly growing business. All of the factors underpin our conviction that 2025 will represent an inflection point for revenue growth for Accelsius. Josh, Dino, and the entire team remain laser-focused on execution and our two market disruptors. Our confidence in the business has never been higher than it is today.

We are excited to watch Accelsius on its journey to become what we believe is at least a $1 billion enterprise value opportunity. With that, I'd like to turn it over to Dave to cover our financials. Dave?

Dave Yablunosky (CFO)

Thanks, Bill, and good afternoon, everyone. Innventure's first quarter revenue was $0.2 million, representing management fees we collected from our management of the Innventus ESG Fund. This was in line with our expectations and consistent with the communication on our last earnings call that we expect most of our revenue growth would be weighted to the second half of the year. As Dino highlighted in his remarks, there is a significant unmet market need for two-phase direct-to-chip liquid cooling, and we believe Accelsius is well-positioned at the forefront of adoption of this technology. We are excited about the momentum that is building at Accelsius.

G&A expenses were approximately $20 million for the quarter. This number has four primary components: non-cash equity-based comp of approximately $5 million, professional service fees of approximately $6 million for accounting, audit, and legal advisory services related to, among other things, regulatory filings and compliance as a new public company, $6 million for payroll expenses, benefits, and other operating expenses primarily driven by growth at Accelsius, and $2 million of non-cash amortization related to our intangible assets. It's important to point out we have deliberately utilized outside professional services to keep our long-term fixed costs related to headcount low. We expect our need to rely on outside professional accounting and legal services to decrease over time as we establish additional internal processes for compliance with public company requirements.

Moving down the income statement, we booked a $233 million non-cash goodwill adjustment driven by the write-down of goodwill on the balance sheet. This was a result of the decrease in the company's share price and market capitalization, which were in part due to the general downward volatility experienced in the stock market during late February and March. Again, this was non-cash. In the non-operating section of the income statement, we booked a favorable non-cash adjustment of approximately $16 million related to the change in fair value of our warrant and earn-out liabilities. EBITDA for the quarter was a loss of approximately $248 million, but after adjusting for non-cash items, adjusted EBITDA was a loss of $21.8 million. To say again, we believe 2025 will represent an inflection point for revenue growth at Innventure.

Due to our majority ownership of Accelsius, we expect to see the results of this revenue and earnings growth in our consolidated financial statements in future quarters. Moving to the balance sheet. As highlighted on our prior earnings call on March 24th, Executive Chairman Mike Otworth, Chief Strategy Officer John Scott, and another related party terminated approximately $18 million worth of Innventure and AeroFlexx debt in exchange for approximately 2.3 million Series C preferred shares. This resulted in an annual interest expense savings of approximately $3 million and reduced our cash expenditure obligations by $18 million. This action underscores our founders' commitment to and confidence in our business strategy. In addition, in the second quarter, pursuant to our securities purchase agreement with Yorkville Advisors, in two tranches, we issued convertible debentures for an aggregate amount of $30 million.

In connection with these issuances, we received gross proceeds of $18 million on April 14 and $9 million on May 15th. This added $27 million of cash to our balance sheet. These two actions illustrate the active management of our cash and both improve our capital position. Other notable items on our balance sheet: inventory. Inventory remained steady from the end of the prior quarter at approximately $5.2 million. We're acting judiciously with our cash, optimizing the need to keep inventory on hand to meet customer orders while keeping cash on our balance sheet. Intangible assets. This represents developed technology and other intangible assets and amortizes at different intervals, primarily over the next eight to 15 years. Goodwill. As mentioned earlier, goodwill now stands at approximately $437 million as of March 31st. On to the cash flow statement.

In the cash from operating activity section, you can see many of the non-cash items I mentioned earlier. The cash used in investing activities primarily represents the funding given to AeroFlexx through our existing facility comprised of debt securities. The cash provided by financing activity section includes approximately $8.2 million of net cash raised through the Innventure Series C preferred stock round, the issuance of shares through the Yorkville standby equity purchase agreement facility, separate from the advance mentioned earlier, and cash raised through Accelsius. It's important to say the Innventure team remains focused on additional capital raises at the Innventure and/or operating company level to meet our liquidity needs and fund what we believe are very attractive growth opportunities ahead of us, with the goal of creating substantial value for our shareholders. With that, we'll open up the call for questions.

Operator (participant)

Thank you.

Ladies and gentlemen, as a reminder to ask the question, please press star one one on your telephone, then wait for your name to be announced. To withdraw your question, please press star one one again. Please stand by while we compile the Q&A roster. Our first question comes from the line of Nehal Chokshi with Northland Capital Markets. Your line is open.

Nehal Chokshi (Senior Research Analyst and Managing Director of Technology)

Thank you. Thank you for bringing Dino onto the call to talk about in detail what Accelsius is doing to capitalize on this massive inflection point for DCH DLC. Going into that discussion there, with respect to the go-to-market, addressing the four different customer types, with the white label agreement with the OEM, was that primarily driven by an end hyperscaler engagement?

Bill Haskell (CEO)

I think it's broader than that. No, I'll answer the question, by the way. I can flip this over to Dino in just a second.

But the OEM that we're doing the white label for has a much broader market than just the hyperscalers. Certainly, they serve various hyperscalers, but they serve a broader market. But Dino, if you have any other commentary on that, feel free to jump in and add on.

Dino Foderaro (CRO)

Yeah. Yeah, Bill, the relationship was there too, driven not only by a hyperscaler, as you mentioned, but more of the overall market demand and kind of where the market is headed. So it could serve as hyperscalers as well as, as you mentioned, a wider breadth of customers.

Bill Haskell (CEO)

Like colo players and other stuff.

Nehal Chokshi (Senior Research Analyst and Managing Director of Technology)

Right. Right. Totally understand that going through this OEM will address the broader swath of customers that that OEM can address. But I guess what I'm trying to get at is, did the evolution or the original relationship start with a hyperscaler?

Bill Haskell (CEO)

Again, Dino, I'll defer that one to you. You're closer to the ground on that one.

Dino Foderaro (CRO)

Yeah. It didn't start with a hyperscaler. It started basically with the OEMs driving kind of where they want the solution set or the product set to go from a liquid cooling standpoint and the goals and objectives they want the infrastructure side to meet. So it was definitely the relationship started from specifications that drove towards a larger, like I said, drove towards the larger market. Now, we do see hyperscalers being able to use this, Neoclouds being able to use this, as well as other users of the data center.

Nehal Chokshi (Senior Research Analyst and Managing Director of Technology)

Okay. And with respect to this potential inflection point happening in calendar 2025, do you expect it to be served largely through this white label agreement?

Bill Haskell (CEO)

I would say broadly, no.

Meaning, while we certainly expect some volume from those agreements, there's a much, much broader pool of customers that we're engaged with. Again, Dino, if you want to amplify on any of that, feel free to jump in. There's a broad range. Our activity is kind of through the roof right now in terms of the activity coming to us from all of those various sources: the hyperscalers, the colos, the as-a-service, and the OEMs. Anything to add, Dino?

Dino Foderaro (CRO)

All right. No, Bill, I think you nailed it there. Yeah. Our pipeline is definitely filling with customers from all angles or opportunities from all angles, especially those being driven by our ecosystem of partners. They've really turned it on. The sales force has really been able to find a lot of opportunities that our product really addresses the need of that customer base.

Not just driven from that OEM agreement, driven from the entire set of relationships that we have out there in the ecosystem.

Bill Haskell (CEO)

I mean, one of the things you saw from the chart, Nahal, is we had six of these partners in January, and now it is 21. All of those have the ability to—they are all channels of one form or another, right? We are getting, I will not say inundated, but there is a huge amount of activity going on in the space. In terms of the trajectory, I think the entire industry was slower in the first and second quarter than they had anticipated. You can sort of see that across the board. I think they were all quite bullish on the second half of the year, and that is what we are seeing.

I think that slowdown in the first half was attributable to a number of factors, but mostly tariffs, uncertainty, and then NVIDIA putting out their roadmap in late February of where they see the chips going when and what type. Since that inflection point, it's been, as you can see in the data, a significant ramp-up in activity.

Nehal Chokshi (Senior Research Analyst and Managing Director of Technology)

Got it. That's great. Thank you for that detail. Dino, could you comment on the pros and cons of flow versus pool-based cooling?

Dino Foderaro (CRO)

Yeah. Bill, do you mind? I can jump in and briefly address this. Yeah. Go ahead. When we look at the pool boiling versus flow boiling, you do put in moving components, potential failure points into your cold plate technology that could cause issues in a long-standing operation, especially as we look at reliability and uptime being extremely important to these customers.

Additionally, when you have pool boiling, you really have no excess for any overload or any hiccup in the facility water supply. You are really exiting at 100% vapor quality, and that could lead to cases of dryout. Whereas the flow boiling, we always have ample fluid flowing through the cold plates and through the system that allow that additional headroom or a little bit of extra buffer. I think a rule of thumb from engineering is never to really design something size on size or for the exact max requirement. When you look at pool boiling, that's really what you're doing because you do not have that excess fluid that's available there if you go outside of normal operating conditions.

Nehal Chokshi (Senior Research Analyst and Managing Director of Technology)

What would you say the pool-based boiling guys would point to as shortcomings for the flow-based cooling?

Dino Foderaro (CRO)

I think the initial point that they point at is pumping power and whatnot and the fact that you have to have a little bit more fluid inside your system. However, when you look at the cost associated with extra fluid, especially with the fluids that we're utilizing, which are commercially available, relatively low-cost fluids when you talk about overall liquid cooling fluids in general, and the pump power being below those thresholds set by the OEMs of that 1.5%, 2%, the benefit that you get from having that resiliency, that robustness in the product that you get with the flow cooling versus the pool boiling, we see that trade-off being a no-brainer going towards the flow side.

Nehal Chokshi (Senior Research Analyst and Managing Director of Technology)

Great. And then last question for me, and then I'll get back into the queue.

My understanding is that there are some hotspots on the Blackwell architecture, which could be a potential accelerant towards two-phase DLC. Any reconnaissance on whether or not NVIDIA can and will mitigate those hotspots on the next-generation architecture Rubin due out sometime in calendar 2026?

Bill Haskell (CEO)

Again, Dino, that is in your territory, not mine.

Dino Foderaro (CRO)

Yeah. I would prefer not to answer that question. I would prefer not to answer that question just due to NDAs and whatnot.

Nehal Chokshi (Senior Research Analyst and Managing Director of Technology)

Gotcha. Okay. Understood. That is perfectly reasonable and understandable. I will cede the floor. Thank you.

Operator (participant)

Thank you. Please stand by for our next question. Our next question comes from the line of Chip Moore with ROTH Capital Partners. Your line is open.

Chip Moore (Managing Director and Senior Research Analyst)

Hey. Thanks for taking the question. Wanted to follow up, I guess, on appreciate the deep dive on Accelsius.

I think the tone around revenue inflection certainly sounds like you're feeling more positive on sort of near-term trajectory. Maybe just expand on that pretty notable increase in lead generation there since February, I guess. You talked about fulfillment being able to be pretty quickly. Just maybe talk about potential to leapfrog single-phase, sort of is better now, needed later type mentality, how those conversations are going, and your confidence level in that inflection point here in the second half.

Bill Haskell (CEO)

Yeah. Let me just start, and then I'll let Dino add on. First, Chip, thanks for the question, and thanks for the interest. We have seen, as you saw in the data, a very, very big spike in activity around late February. From that point forward, it's been significant. That data is, I believe it's kind of a three-month moving average.

When you see it triple from month to month, you can see a pretty significant explosion in activity there. I think that parallels what we're seeing from all the other larger players in the space. I listened to some of the other earnings calls and some of the other principals in the marketplace, and I think they're experiencing the very same kind of activity. We're kind of aligned with that. Obviously, we can't give you any kind of forward-looking guidance on revenues. We're not prepared to do that. In terms of the momentum, if you want to kind of drill down with Dino on the momentum, that's perfectly good. Absolutely. Bill, if you have anything in particular you want to add on to that.

Dino Foderaro (CRO)

Yeah.

In regards to momentum, I think we touched on it in the readout and the update, is that with the, I guess, clarity provided by NVIDIA's roadmap that was released at GTC and then released even further at some of the conferences following that, really gave direction to the data center operators, the infrastructure manufacturers, and those that are kind of in the know on where the processes are going, on what rack architectures will look like, really started to draw attention to the extremely high flow rates, not only at the rack level, but also at the chip level for single-phase cooling. That really drove a lot of attention towards two-phase. Additionally, there was a very small amount of liquid cooling deployments that had actually been achieved in the prior years outside of the hyperscalers.

As enterprise clients, as colos started to deploy more of the single-phase water, those problems, those issues that are inherent to that technology really came to roost, and people got to feel those personally. I feel that drove a large amount of attention and really kind of opened up the market to what we've been evangelizing over the last few years. That has definitely resulted in that drive of interest and the drive of right interest. There has been a big paradigm shift in the leads that we've captured and the conversations that we've had where it's no longer, "Oh, this is a neat technology.

I've been saying that two-phase is going to be needed to the people that are the experts, the subject matter experts at the colos, the OEMs, the people that are actually deploying the cooling, the MEP firms, the design firms, to how do I deploy this? Your solution is viable. What is the way that we get this into our data centers? That is really, like I said, the drive and the paradigm shift, the types of conversations that we've had has definitely kind of pointed at that inflection point and that wave coming out of Accelsius.

Chip Moore (Managing Director and Senior Research Analyst)

Great. That's helpful. Appreciate it. Maybe just one more follow-up, and I'll hop back into you. Just maybe talk about that sort of chicken and egg confidence in your ability to scale and manufacturing, how those conversations have progressed. Thanks.

Dino Foderaro (CRO)

Yeah.

Bill Haskell (CEO)

Go ahead, Dino.

Dino Foderaro (CRO)

Yeah.

I mean, as we look at the manufacturing side, obviously, we've got a great team. One of the first people that was added to the organization when we were founded is our Chief Supply Chain Officer, tremendous amount of background there on the supply chain side. He's added a great team. Myself, being a manufacturing guy at heart and running manufacturing companies, we knew that we were going to need the support to add confidence for the hyperscalers to large clients. That's where our CM partner and our CM partners have really come in to drive a lot of that confidence and help ensure that we can support those large volumes, as well as ensuring that those volumes can be met with quality that's expected in a tech market or a high-reliability market.

That was really a big add for us in overall ensuring our clients that we could grow with them because nothing's worse than testing, evaluating, and adopting a technology and then finding out that it can't scale with your demands. That's kind of where we were able to add a lot of confidence from the manufacturing side. Now, when we look at the design development side, again, great team of engineers and a great sourcing group that was able to identify the proper materials and really a robust process to ensuring that the materials that we utilized can meet that high-reliability requirements of our clients, which I'm not sure a lot of the other units that are out there or new technologies are out there take to heart because they're more concerned about getting the technology out.

The 360 approach that we took, I think, has lent very good confidence or gained very strong confidence from our clients.

Bill Haskell (CEO)

Yeah. I think another way to frame it, Chip, is that, as we mentioned, as Dino mentioned, we've had a lot of engagement with some of the hyperscalers that have very high volume demand, and they're confident based on the interactions they've had with us, or they wouldn't be coming back. The fact that we've had, in some cases, a dozen meetings or more with various large hyperscalers suggests that they see this as a viable path to getting to scale. Of course, they understand that we're partnering with content manufacturers to handle the larger volumes. Even internally, with the kind of volumes we talked about, we can still get to some pretty meaningful size just with our current internal manufacturing capacity.

It is almost exclusively a kind of North American supply chain, at least our direct suppliers. That was purposeful to kind of take out the geopolitical risk out of the equation so that we would be able to access the materials that we needed in order to accommodate the volumes that we can manufacture internally.

Operator (participant)

Thank you. Ladies and gentlemen, at this time, I would like to turn the call back over to Bill Haskell for closing remarks.

Bill Haskell (CEO)

Thank you, everyone. I really appreciate everybody joining today. It is useful to get a line directly from the horse's mouth, which is why we wanted Dino to come on to talk more in detail about Accelsius. We do look forward to future updates. We believe that there is, again, a lot of momentum going on in the company, in all of our operating companies, quite honestly.

I feel that we're very well positioned, and they're all really maturing and kind of coming to the fore. Very optimistic about where we go from here, but we look forward to updating you in the future. Thank you again.

Operator (participant)

Ladies and gentlemen, that concludes today's conference call. Thank you for your participation. You may now disconnect.