This call and all of Water Tower Research is available to everyone. It can be found on our website, and a recording of this chat will be on our website soon after this call. We will be taking questions from the audience, and those can be submitted via the webcast. Hopefully, we'll be able to get to all of them, but if not, our apologies if time does not allow. With that, John, welcome back. What's new at Soluna? I think it was late January we had you on last.
Graham, thanks for having me back. Always a pleasure. I enjoy these fireside chats. Yeah, I think the last time we spoke was in January. We had just published our report on the lower carbon footprint of our flagship project, Project Dorothy. And since then, a lot has happened. I feel like a year has passed. We launched the first AI pilot at our Project Sophie facility. It's a young startup that is focused on basically delivering AI solutions for customers. Excuse me, they're specifically focused on the AMD platform, which is interesting. We did some modifications to the convertible note or debt agreements that we have, and that was focused on bolstering our balance sheet and restructuring the note so we can continue to reduce that balance. We announced our Q4 results and our annual results. Folks can find that out there. Very excited about that.
We published record revenues. We were up to $10 million compared to the third quarter revenue of $5.8 million, so a 75% increase. Our gross profit was up. It grew to almost $4.3 million. It's probably the highest profit quarter we've had since the original MTI company started the crypto segment in Q2 of 2020. We had another consecutive EBITDA positive quarter as well. Stronger balance sheet. Ramp on Dorothy. Dorothy 1A, 1B ramped combined $13.7 million in revenue. And the comment I think I made on the release was that we were trying to keep our boat afloat while rebuilding the boat and focusing on a new direction. And so there was a lot of rearchitecture of our business that took place last year, and that was designed to improve our growth potential, revenue, our hosting business, and also add diversification to our revenue.
So we added ancillary services, and now we're venturing into this AI business. So a lot has happened, Graham. I think I've gotten everything. Dorothy 2, I forgot. Actually, there's more. Dorothy 2, clear modeling. We actually started focusing all of our energy now on the construction of Dorothy 2, and folks will be hearing more about our financing plan for Dorothy 2 in the coming weeks. The Kati PPA, we've been working on that as well, and the ERCOT process for that as well, and we're in the final stages of that. So there's just been a lot. This happened since January, and I feel like only a short period of time has passed, but a lot has happened.
Yeah, things are definitely moving ahead for you, and you guys have come an amazingly long way since a year ago in terms of keeping the company afloat and then getting the ship in the right direction. Do you want to dig in a little bit more to the AI part? Can you talk about what you're doing on the AI front, what you're doing at Sophie, and then how does that compare different from what you're going to be doing at Dorothy 2?
Yeah, sure. So taking a step back, we've long said that we have a strategy beyond Bitcoin. And in the last year, as everyone knows, I don't think you can escape it at this point, AI has become a major momentum and technology transformation that everyone is talking about. And we saw that as an opportunity now that we had sort of finished up all of our cleanup and getting our projects up and going. We can now focus our energy on taking advantage of that opportunity. And what we're doing is sort of a two-step approach. Number one, at Project Sophie, we're reserving some capacity for companies that want to do the sort of classic colocation hosting type of relationship. And we're focusing on companies that have a sustainability focus, want to partner with us.
And there, what we've done is completed the installation of GPUs for our first colocation partnership. That's that startup I talked about. They're specializing in AI computing services. We're working pretty closely with them, and they seem to be growing very fast. They're starting to get a lot of traction, and so we're expecting to see some continued growth with that relationship. It's also helping us to learn about the requirements for customers that are in that business, the infrastructure improvements that we would need to make to Sophie, and we've already made some and planning to make more, which will be designed to make that facility more conducive to AI-centric data center operations. The second thing we're focused on is Helix. Helix is a new design. Folks are familiar now with our sort of diamond-shaped design, which we now have a patent for, by the way.
So the configuration of our facilities for thermodynamics management is a patent we've won, one of many that we've filed that we hope to receive. Helix is a whole new initiative to take all of our learnings around AI and our experience with Dorothy 1 and build a whole new data center design specifically tuned to AI workloads. AI workloads are very power-dense. In fact, they're 10 times as power-dense as your typical compute environment. And what we're going to do is we're going to build the first instantiation of Helix at Dorothy 2. So Dorothy 2 was originally planned to be a 50-MW Bitcoin facility. We're now making it a 48-MW facility, and we're going to place at least 2 MW of this new Helix-type facility there.
Our plans are to not only build out that facility, and by the way, the difference in that design, it's a sealed building. It has the capability to do direct liquid cooling on the racks that the GPUs will be on. The building will have a special HVAC design. It'll actually have a very special configuration that allows it to scale just like our traditional data centers, but it'll be larger and more of an enterprise-class facility. We'll also be pulling new connectivity to Dorothy 2, which will make it, again, more conducive to this type of compute. We're working with a number of different partners to help us to vet that design, make sure we've thought about all of the things that we'll need to understand. We've laid out different racking configurations and racking technology, different compute.
We're also thinking about that next stage of compute that Jensen Huang at this point is like Michael Jackson. He's like the Michael Jackson of NVIDIA of AI at this point. And at his recent concert in San Jose, where we attended, he talked about the next rev, right, going from Hopper to Blackwell. And that's going to be sort of the next set of capabilities that are going to be deployed out there. So we're anticipating that. So the goal is to build this facility at D2. And in fact, it's on its own fundraising path. So our Bitcoin business is continuing to grow and expand, and we have a capital strategy for that. And then now you're going to start to see a capital strategy formation happen around AI and specifically Helix and some of the other initiatives that we have going on.
There's a lot going on there. We haven't put out to you folks and to the market the full AI strategy and initiative, but that's coming here soon where we'll start to unveil some of the partnerships that we have and the approach and Helix and all.
What are you hearing from customers or potential customers about your solution? If there's any feedback you could share.
What can I share? Power, power, power. I think a big driver. Dip and I spent a week at the NVIDIA conference meeting customers, partners. We're now beginning to talk to the NVIDIA cloud program folks over there. And what we're hearing is that the biggest challenge facing the growth of AI is power consumption. And renewable energy is really ready to scale. Look, it's more affordable, especially the way we're doing it, where we're behind the meter. We can get access to large amounts of energy in a short period of time, relatively speaking. Right now, we've heard tell that there's probably a two to three year window where there's really almost no source for new power development and new sites that can access power. That's a major problem for the space. That's probably why you're hearing a lot of talk around nuclear, around new approaches to power.
That's being driven by the fact that AI has a real sustainability issue. It's power-hungry, it's thirsty, and it's dirty, depending on where it's based. Customers are looking for a sustainable platform that they can start testing their AI projects and then scaling that. They need something that's sustainable and scalable. More importantly, they need a full stack. A lot of providers out there provide just bare metal or access to computing power. We want to take that one step further, and that's why we're working on some very special partnerships that will allow us to build really a sort of turnkey solution for businesses that don't have the expertise to go after AI. This is leveraging some of my enterprise software muscle, if you will. Reminder, I really started my career as an enterprise software entrepreneur.
I understand that businesses always want to leverage the next generation of technology. To do that, they don't always have the expertise to do that leveraging. They want a partner that can help deliver them a turnkey solution. For AI, what that means is I want to be able to download a model that can be useful to me, tune that model, and then deploy it in my enterprise to help me to scale. We're going to be focusing on delivering that full solution, a sustainable, scalable, turnkey solution for generative AI.
Great. Yeah, no, the power consumption is going to be a challenge in this country where the world's going to face. And certainly, if we want to move to more renewables, it has to come from somewhere. So unfortunately, right now, the solution is keeping coal plants online longer. So that's where the more renewables can come out, the more that can help. So that's where you guys have an important role in that solution. One of the questions has come in, paraphrasing it, but basically, can you give us an update on your pipeline in terms of what you've got in front of you and any timing around that?
Yeah, so to start out, when we reference our pipeline, we always talk about the near-term shovel-ready projects and then the longer-term pipeline of developmental projects. The two that we've been focused on for the last year or so have been the next phase of Dorothy, so Dorothy 2. And that's the next 48 MW, if you will, of our Bitcoin hosting business. Our power partner's approval process is near complete. This is where we tell them, "Hey, we want to double the site that we've got back there behind the meter." They take that to their sponsors, get approval, and so forth. And so we show them the construction plans, the timeline, and how we're going to connect to their substation. So in the process of doing that, we submitted to ERCOT the model for our facility.
So that's now in the ERCOT model, allowing us to plan an energization strategy once the construction reaches that point. We kicked off a bidding process earlier this year. We identified some of the best contractors in Texas with experience in high voltage and civil and electrical to help us to build out the next phase of Dorothy. Dorothy 2 is a cookie-cutter of Dorothy 1. So there's not a major sort of new risks associated with the construction. But we wanted to upgrade the partners that we were working with, use the best possible contractors, and learn from our experience in Dorothy 1. And so we're about to make some awards here of who will be the winners for the process and then start to kick off the notice to proceed to start the construction. Kati is a 166-MW project we've been talking about.
Kati or Kati is named after Kati Karikó, the inventor of mRNA, if you will, the researcher that really led to those mRNA vaccine therapies that we benefited from. And sometimes on some podcasts, she calls herself Kati. So we sort of go in between, but we like calling them by what the person's name is. So it's Project Kati. That project is in the final stages of commercial agreement on the PPA. I've been directly involved now that they're reaching maturity. The documents are pretty solid. Our partner is a very experienced partner, huge. So they have a much bigger team working on this. So things have gone much quicker than Dorothy 1. And that is based on the fact that they are really focused on this as a very strategic project for them. The ERCOT planning process has also moved along quite quickly.
We're in the final step of the planning process. There's a specific study called the reactive study. We've gone through that, and they've given us some final feedback on that, as I understand it. The power partner is the one involved in sort of fielding those. We're going through that. We have the landowner negotiations that we're doing now. Securing the land where the project will be based and selecting the specific site, we're working on that and getting a lease agreement that will ultimately be satisfactory to both sides or three sides, rather. I just met with one of the main suppliers we're looking at for critical equipment, like transformers and power equipment that we'll need for this site that will be pretty sizable. There's a lot happening.
I just want to say to folks, when they see us put out our monthlies where we put out our metrics and information about our different projects, and Kati just seems to sort of stand still. Just remember, this is a development project. It's moving actually quite fast. But when you shrink it down to monthly scale, not a lot happens, right? But we are making progress. And then our larger pipeline is fantastic, Graham. We're getting lots of folks calling us inbound now. When we first started, about, I'd say, 70% of our pipeline growth was driven from us calling out to companies. Now, folks call us, and these are some of the largest infrastructure folks in the world. And so we've got a couple of other projects that are coming down the pipe here that will start their accelerated development process as well. So lots of progress there.
Great. Thank you. In your last monthly update, you highlighted the curtailed energy that you were using. You had a big bump-up month-over-month. How does that benefit, I guess, what's driving that bump-up in curtailed energy? And then how does that benefit flow through to both you and the wind farm?
Oh, it's a great question. So as a reminder, curtailed energy is energy that the wind farm can produce. The wind is blowing, the sun is shining, but it can't monetize that energy. Usually, it's because the grid is overwhelmed with power options, if you will. Depending on where you are in the grid, there might be lots of wind farms or lots of solar farms. And depending on the time of day, there might be lots of energy being produced, but not as much load to consume that energy. And so you have curtailments that are either technical in nature, where the power line just can't deliver all of the power, so congestion, or what you have is basically a capacity problem where it doesn't need all the power. And so those curtailments are signaled to the power plant with negative energy prices, typically.
When power is negative, we will consume that power, and we pay the power plant a positive price for that energy. In February, we saw a huge increase in the amount of curtailment at the wind farm, at our Project Dorothy. That has a lot to do with seasonality. A lot of times I'm sorry, annually, wind gets really serious down there. In fact, you can be down there, and it can be a pretty warm day, and then you drive to the site, and you're freezing to death because there's so much wind just gushing through there. That excess wind generates a lot of energy. If the grid can't support it, it becomes curtailed. What we do is we ramp up the facility and consume that energy and pay a positive fixed price for that power.
What the wind farm benefits from that is they get all of that excess revenue, that extra revenue that would have been zero or negative for them. In other words, they would come out of pocket to send that energy. That unlocks their production tax credits, their renewable energy credits, and so forth that are all beneficial to all the constituencies there. I think in February, if I remember correctly, we had somewhere around 11.9 or 11,900 MWh of curtailed energy. That's almost a doubling, more than a doubling increase since January. Needless to say, it's very windy down there. What happens is things change. In the summer, it gets less windy, so you'll see less curtailments. It's very hot down there, and the wind tends to die down.
A lot of that curtailment is driven by the amount of energy that's in the area and the cyclicality driven by weather patterns down there.
Add to that, there's more demand with air conditioning on the grid, but then also the wind typically blows a lot during the middle of the night where you have your lowest power demand, so.
Yeah, that's exactly right. And the beautiful thing about that, putting it back in the context of AI, is you have pockets like this all over the country that are fantastic resources for power generation, but that power generation doesn't have a natural home or natural sink for that energy. And alternatives are batteries and transmission. And we've talked at length about the fact that these aren't scalable solutions that are available everywhere today. But AI computing, you can move that computing to pretty much any location, right? You don't really care where the results from your ChatGPT prompt is coming from.
You actually don't know.
You actually don't know, right? I don't even think people think to wonder where it might come from. And so the idea is, since computing is inherently distributable, you could distribute it out to these locations where you have stranded power. And if you can develop technology like we are doing to convert that power with high levels of efficiency to as much computing as possible, and you have software to manage the uptime and support the source of energy and so forth, you can deliver a distributed compute fabric that can be used for a host of different batchable computing applications like AI and Bitcoin mining. So it's pretty exciting.
I had a question come in around the halving of Bitcoin. And I know that in your recent Ask Me Anything that's posted on your website, you talked about that a little bit and saying that it tends to push out the less economic miners. Do you see any impact to Soluna from this?
So look, the halving is there's lots of questions about what's going to happen in this halving. In the past, halvings have been catastrophic for miners. So miners have literally gone out of business. The industry has matured where there's planning for halving over a year in advance before it happens. And most of that planning is focusing on the efficiency mixture of the machines, making sure you have the best possible machines, the best power pricing, the best location for your miners. You're scaling as a company and so forth, and a host of other optimizations that are now available to the market, like software and things like that. And what's exciting at Soluna, we've done all of that. In fact, by design, we're like the perfect home for miners looking to place their machines as they scale through into and through the halving.
So if you look at our facilities and our business model right now, more than two-thirds of our business is hosting. And if you look at the elements of our business that are directly exposed to mining, in other words, proprietary mining, that's Dorothy 1B and Sophie. We've got some JVs there. And so our average efficiency at Dorothy 1B is about 28 J/TH. And across our fleet, it's actually less than that. I think we calculated recently about 25.7 J/TH. That's by design. We actually don't let people into our facilities unless they have efficiencies within ranges that we think will allow us to continue to be partners through the halving. And sometimes we get questions, why are the hosting agreements you sign so short? We sign them two years, sometimes 1 year in cases.
That's so we can checkpoint at the end of those terms and look at the efficiency of those miners and either ask the partner to upgrade them or replace them with a partner to do that. Toward the tail end of last year and the beginning of this year, we have been reconfiguring or bringing in new partners that increases the profitability and structuring within Project Sophie, which is our hosting plus profit share. Of course, 1B, we bought the most efficient machines we could get at the time. Then when you combine that with our power price, which is sub-$30 per MWh or sub-$0.30, that gives us a meaningful cost advantage that is very important relative to many other miners post-halving.
So with our partner, Navitas Global, we have an approach that allows us to look at the equipment strategy and potential upgrades to the extent that makes sense. And then Project Dorothy 1A and Sophie, the partners that we have in there are some of the largest miners, and they have some of the most efficient fleets in the space. So the short answer to your question, I guess, is we don't see the halving affecting us nearly as much as some others out there. But it all remains to be seen. This halving is a bit different from the past thus far. We have seen higher interest rates. We have seen several Bitcoin ETFs that have amassed over $10 billion worth of Bitcoin in inflows from institutional investors. And Bitcoin, as of late, continues to be and remain at all-time highs.
You're also starting to see transaction fees really spike up because of new applications being built on the platform. This will be an interesting having to see over the next six months. Generally speaking, the industry is pretty positive about what's happening. In any case, we've made all the preparations that we can at this point to prepare us for it.
Great. Thank you.
Oh, and one more point, as I said earlier, we added cash to our balance sheet also to protect us, just like everybody else, so.
Always good position to be in.
Yeah, exactly.
Coming up towards the end of our time, so I just want to ask a final sort of wrap-up question. What are investors missing, or rather, what should investors be focusing on as we move forward from here?
Yeah, thanks, Graham. I get this question a lot. This is my thought process, Graham. Since I've been in the role I think it's like 18 months now. I'm losing track of time.
Not even that. It's like.
Is it? Is it a year?
Yeah, what? It was like May of yeah, because I think.
Wow.
It was about May because I know you kind of sort of took over in March, and then.
Yeah, you're right. You're right.
It probably feels longer to you given what you've been going through, so.
You're right.
Like 20 years already.
You're right. Well, thanks for helping me put that in perspective. A lot has happened that time. I think about the approach that I've taken with our shareholders, and I think I've tried to be as transparent as possible, is one of our core tenets inside the company, and we project that outside the company too. So we put out monthlies. We do AMAs. Thankfully, we spend time with you guys to share about our company. So there shouldn't be much that people don't understand about the business. But I did give some thought about one thing that I think people are missing. That is, in a word, it's resilience. To help you understand what I mean, I want to tell you a bit of a story. So Friday, just like everyone else, I had this sort of tracker on the 840,000th block.
So everyone knows, late Friday afternoon, that block was triggered and added to the blockchain. And it started the 5th epoch of what has become the 15-year experiment of Bitcoin, right? And that triggered for me a sort of flashback of where were we as a company 4 years ago? At that time, if you remember, Graham, it was 2020. So we were all locked up in our rooms experiencing the world's 10th pandemic. Soluna was still working on proving our business model, and we were working to actually win the bid for our first data center, Edith. So that was our first instantiation of our project. It was a much smaller facility, about 3 MW. Today, fast forward 4 years, we are a strong brand in the renewable energy curtailment solution space. We have built two iterations of our proprietary data center.
We've proven our software technology, MaestroOS. We've grown a team and a culture. We've survived what many should agree was a catastrophic collapse in Bitcoin price. We have proven our business model. We've amassed over 2 GW in our long-term pipeline, which is up from 5 MW four years ago. Think about it, from 5 MW to 2 GW of pipeline. And now we're venturing into a new frontier with AI. With this retrospective, that sort of quick moment that I took to think about the business was that it's our resilience. Our resilience is underrated, Graham. There's this concept called the Lindy effect, and it's basically a concept or a rule of thumb that predicts the future life of something. So if something's been around, it's more likely to continue to be around.
The Lindy effect is used to look at technology and ideas and everything's sort of within the current age. It just suggests that the longer something is here in proportion to its future life, that increases over time. We're like a poster child of the Lindy effect. In an interesting parallel, so is Bitcoin. Its resilience is also underrated. Every time it seems like the latest explosion in the ecosystem is the final blow, and it just rises again, even stronger. The technology is now going through an interesting maturity phase, right? I guess more importantly, its strength leads to more maturity, and it therefore has more staying power. In this way, Soluna and Bitcoin are kind of kind of kindred spirits to some extent. I'll leave our shareholders out there and folks who are looking at our company.
What folks are missing is that we're an extremely resilient team. We have a clear vision. We have incredible resolve. We have a strong ability to execute. We are now focused on the growth of the business. So I want to thank all the shareholders and investors who can see this and have been patient with us. And for the rest of you, stay tuned.
Well, great, John. Thank you so much. Appreciate your time today. And thank you, everyone, for joining us. To learn more about Soluna, please visit their website or access our research on the Water Tower website. Also, I'd like to point out the views expressed in this fireside chat may not necessarily reflect the views of Water Tower Research, LLC, and are provided for informational purposes only. This fireside chat may not be distributed or reproduced without the written consent of Water Tower Research and should not be considered research nor a recommendation. Water Tower Research is an investment management firm, not a licensed broker, broker-dealer, market maker, investment bank underwriter, or investment advisor. Additional disclaimers can be found at watertowerresearch.com. With that, thanks so much, everyone.