DUG Technology Ltd (DUG) Earnings Call Transcript & Summary

Welcome, ladies and gentlemen, to DUG's interim financial statements for the FY '25 year. Today, we have Matt Lamont, Founder and Managing Director; and Daniel Lamont, who's acting CFO, joining us. They're going to take us through the financials. Just a bit of housekeeping. Research analysts are free to ask questions at the end. You can do so via the Q&A function or by raising your hand. You can ask them live if you can ask those along -- as you go along the way. Handing over to Dan and Matt.

Thanks, Steve. Let me share the presentation. It's my pleasure to be with you again today with Dan. And I think it's exciting times for DUG, although the results are not quite what everyone would have hoped for. We're really excited. So let's get on with it. So the highlights of the year is the order book. It's now growing again, and it's growing on the back of the MP-FWI, very strong forward momentum. The software revenue is up 22%. I've been sort of talking about that over the last few releases saying that we're really excited about software. And finally, we're starting to see that deliver, and I think we'll continue to see that accelerate. Revenue was down a little bit as we flagged at 6 months ago, we were going to be flat. And unfortunately, we ended up slightly down on flat. But -- and EBITDA was soft. But we've been investing a lot because it's sort of like being 2 worlds for us, very disappointing on the share price side, but internally, everybody can see what's happening. Everybody knows our technology and can see our customers and our excitement and everything that's going on. And so it's been a funny 6 months in that regard in such excitement and drive and everyone facing forward internally, but this uncertainty in the share market. And the Elastic MP-FWI projects do look really good. We have 8 that are underway, a couple that are nearly finished, and we've had follow-on awards now. So it's going great. It really is. Just rethinking who we are and how we get that message across. Again, we are numerical physicists. We're global -- we work for those global technology leaders in several different ways. We've made breakthroughs in geoscientific computing, in algorithms like Elastic Multi-parameter FWI, first in the world by far. We work on the most complex of data. We work geographically all over the world. We've got some of the best supercomputing facilities and some of the most energy-efficient cooling systems for data. To sell -- to maximize our growth, we have to sell across the entire planet. And so we're now open at Abu Dhabi. So we're now on 5 continents, and we're committed to delivering what we do globally. We're global to our core. So just as a reminder, services is where we do turnkey projects for mostly oil and gas companies, but also for carbon capture and storage, for wind farms and for several other things. But most of the -- the majority of the revenue comes from oil and gas. And that's where the Elastic MP-FWI revenue mostly comes from. It does come from software as well. The software is in 40-odd countries. It is around big data analysis, big data storage, big data mining. And yes, we're really proud of the software. It's a modern, broad software suite that sets it apart in the industry. HPC-as-a-Service, although in its own right, is disappointing, and we've been relying on just a few clients. And as they change internally, they haven't -- the one we lost that affected the results in oil and gas didn't go anywhere else. They just changed their policy on what work they were doing internally. But it supports and is a key part of the whole software story as well. Multi-Client, this is a change. We've been doing multi-client for a number of years, a long time. We've worked for 12 multi-client companies. The change now is just that we will have our own multi-client assets. We've taken risk on multi-client projects before, but now we will have our own multi-client assets ourselves. It's a really big market and a really profitable market and one perhaps we should have changed our strategy in a little while back. DUG Nomad, we're really excited by it. We haven't sold any, but we're very excited. It is genuinely mobile data centers or genuine edge data centers. It's different to -- what's it different to, Dan?

Modular.

It's different to modular data centers. Modular data centers is what you'll see most everywhere else, like dongles. You take them to a mine site and you get a builder in and he sets them up on stumps and then you get a plumber in, you get electrician in and so on and so forth. And ultimately, 1 month later, you have a data center that's ready to go, and it's very good and useful and a good thing to do. Ours are different. Ours, you can turn on while they're on the back of the truck if you've got the power. You can sling it under helicopter turned on, you can put it in the back of a Hercules turned on. It uses our immersion cooling. It's very cool, but it is differentiated. It isn't the same thing. It's genuinely mobile. And then, of course, there's the BAC licensing of our patent. And there's a really cool movie that they put out about how that fits into what they do, and we'll highlight that link to you a bit later on. You've seen this slide before. It's our geographic spread. So you can see orange of the countries we've worked in. So we work most -- everywhere. They're genuinely global. And then there's our offices and our compute facilities. Of course, the new one on the block is Abu Dhabi and then our dotted line is our global fiber, where we transport data and access to supercomputers and so forth around the globe. One of the highlights, I sort of touched on it a little bit in the excitement that is within DUG because we can see the position that we're in and our guys putting those interactions with clients. And because of that and because we believe that we can have a period of sustained growth, we've been investing strongly for quite a while now, very strongly in the Elastic Multi-parameter FWI pilot projects. They've taken a lot. We rolled out a big chunk of technology. And initially, you get little issues and things you didn't expect and you work your way through that, which is why the pilot projects are important. But the results have been really great. They're not -- I think I expected incremental results for 1 or 2 reasons. We're not seeing incremental results. We're getting quite startlingly better results. Middle East is important. We're getting access to some really big opportunities there now. And I think that's going to grow into a pillar in the DUG world. Again, if you see what -- you have the aspirations that we have for sustained growth and the belief that we have that we can do it and become a very significant player in the market, then you need the business development to support that. And that's what we've been putting in place. You need the conferences, you need the travel, you need the interactions. And so we've been investing in that strongly. Our R&D team is exceptional. We don't lead the world in technology and license [indiscernible] and unless your R&D team is exceptional. And so it's made up of 1%, what I like to call 1%, and they're hard to hire. We've had a bit of a really great year last year. They're hard to hire. They're slow to hire. But we actually hired 5 last year, which is quite incredible. That's a fabulous year. And then in infrastructure, we raised money, as you're aware, to do the infrastructure upgrade. But an infrastructure upgrade was already underway as well. That is now complete, and that gives us 20% headroom. And the -- I think we'll talk about the other -- the further infrastructure to finish off the rest of the room is now underway. Well, the planning for it is underway. The goalpost moved a little bit. The new computers are just so energy rich. They're using so much power that they produced a lot of heat. And so we're having to up the ante on how much power we're providing to that new part of the data center and how much heat rejection we're putting into that new part of the data center. I'll pass now to Dan to do the financial slides.

Perfect. Thanks, Matt. So I think we touched on revenue at the top. So revenue was down 4%. We had services down 3% through the half. But we had a really good result through January with $13.4 million of awards, giving us a closing order book of $42.2 million at 31 January. And this really sets us up for a strong FY '25 half 2. Software revenue was up 22%, which Matt touched on was a great result for the team and driven by customer renewal rates, success licensing our processing and imaging toolkit, which is something that's relatively new for us as a business and the increased adoption in the offshore wind market, which is companies using our software to decide where they're going to place offshore wind turbine pylons. HPC-as-a-Service revenue fell 46% on the half, which as Matt touched on, was a disappointing result, but it's really important to highlight that the HPC-as-a-Service offering really helps us drive software sales. So the 2 together really interlock and work well for us as an overall offering to the market. We had employee benefits increased by 10% over the -- in the period. So a large part of this was due to the hiring of the technical team in the Middle East, which occurred mainly through July and August 2024. And so those costs were carried through the period along with some relocation and training costs. We also undertook a restructuring initiative during the half, relocating shared services from Australia to Malaysia and rightsizing some other teams. So that was a USD 0.7 million cost to us in the half with annualized savings of USD 0.9 million moving forward. We had other expenses fall by 7%. So important in this fall was no third-party compute. So over calendar year '24, we've had the arrival of our new Intel and then our new AMD computers, including some that arrived in July 2024. And this has given us great capacity both to pursue the pilot projects that we discussed on, as well as to remove the need for third-party compute. We see these new computers come through in the depreciation and finance expense lines with the step-up there being associated with the new hardware and then the associated asset financing that goes with it. So at 31 December, we had $17.3 million of cash on hand and $24.3 million in debt. So this debt is predominantly asset financing, but contains some other smaller amounts from -- as well. So that left us with net debt of $7 million for the period. In October 2024, we had the successful capital raising, so we raised AUD 31.4 million. So just giving a brief update on how those projects are going and the use of funds. So we raised AUD 19 million for the data center infrastructure upgrades that Matt referenced earlier. So these upgrades are progressing through design and planning stage at the moment. We've had AUD 6.5 million for the Middle East expansion. So we've hired the team there. We've found the office and the fit-out is underway now. So it's working and moving along well. And DUG Nomad, we raised $3 million for. So we've had an early sales and engineering teams have been put in place. We've established the Malaysian supply chain, and we've also attended a lot of really key industry conferences there. So in the balance sheet, one thing -- one movement in the period is contract assets. And so what we're seeing as we move into new jurisdictions and we work with different organizations is we have less of an ability to negotiate payment terms and payment cycles. And so what we're seeing now is a lot of our new contracts are coming through as milestones rather than monthly progress billing. So what we had at 31 December was just a few of these milestones lining up such that it caused a material uptick in the contract assets at the end of the calendar year. So most of these material milestones are due to be unwound in the first quarter of this calendar year. We also have PP&E that's flat over the period. And it's just worth highlighting there that we had the last delivery of AMD nodes in July 2024. So while we had some depreciation that come through the books, that number is flat due to those machines arriving. Turning to cash flow. Payments to suppliers rose and it's mainly a timing thing as opposed to anything else. So the third-party compute that we utilized over FY '24, we were able to negotiate payment terms that stretched over a 12-month period. So we had the last of those payments come through in FY '25 half 1, albeit that we used the third-party compute in FY '24. We also had $6 million of cash invested over the period. And so most of this is related to the final delivery of the AMD notes that I referenced earlier in July '24 and then also the first round of infrastructure upgrades that Matt referenced. So, then the final big item through financing is the $19.7 million, which I touched on earlier, which is the USD equivalent net of fees that was raised in that October 2024 capital raising.

So growth drivers. What's going to drive us forward? Why are we also excited internally? First of all, the ecosystem in DUG, the integrated solutions is going to underpin everything that we're doing. So it's the proprietary, immersion-cooled data centers, right? We save 51% of the power bill, great for environmental, but really great for the back pocket as well. It's the science. It's the modern geoscience software underpins everything we do with all its workflows and analytics and leading algorithms and an expanding multi-client library, and I'm really excited about where that's going as well. For those that don't know Multi-Client, it's where we build up an asset and then we can sell it multiple times to multiple clients. We only do this in a very risk-adverse way. So we won't be taking on liabilities there. And then, of course, underpinning everything is Nomad to get us into places we don't want to build data centers, but you can't take data out of the country, and the immersion cooling. I put out a video recently on the Elastic MP-FWI. And I would encourage everybody, if you haven't had it, to have a look at it. You might want to have a look at it in pieces, and I think we're going to put it out again in pieces because you probably want to just take a piece at a time and understand it. But it walks you through what we're doing and what that competitive landscape looks like and why it's so important both to DUG, but to the industry as well as what we -- of what we're doing. And the key, of course, about this is just how much a well costs to drill in the oil and gas industry. And so expensive geophysics is well worth the investment and getting better results and getting them sooner is really valuable. And that's the key to MP-FWI in general. It replaces traditional processing imaging, provides better results and provides it in a much shorter calendar time frame, less manpower, more compute. And so the results that we're seeing from the Elastic MP-FWI, as I've said, are really something. And we're going to -- it's going to be big news. We're going to see big things coming from this. If you look at the pictures on the right, I'm not going to get into it, but I'd just highlight the right-hand side, sort of 1/4 of the way through -- 1/4 of the way down, if you put your imagination in place, you can see a reef. It's actually a buried reef type structure there. And you can see at the top picture, you've got this sort of quite rough top to that reef. And at the bottom, the top of that reef is quite resolved. I mean that's significant. And that's what you're seeing. That's just an obvious one, but that's what we're seeing throughout the Elastic results. We're getting it cleaned up. We're getting it more resolution and we're getting it better. The oil companies are excited. So the growth driver. As we've been saying, it is the next-generation size imaging. Leading the way is the Elastic, but also the Acoustic as well, terribly exciting. Those pilot projects are going well. As I said, you're rolling out a big chunk of technology, so you have little issues, which is why you do the pilot projects. We're cleaning those up. The results look tremendous, and we're already getting follow-on contract awards on the back of those. So that's part of why you sort of saw flat awards last year and then suddenly you're starting to see momentum build this year. It's around some of those results starting to come through. And Multi-Client, as I said, it's a high-margin repeat license business, and it can bring great returns. And we've got -- we're looking at several really great projects. Now it can be risky, but we're doing it in a very non-risky way and that we need every project to be underwritten before we do it. And so we're not taking a risk. The software. I've been talking about software for 1 year or 2 now, and we're finally starting to see the numbers come through. I've been saying that we've got a new lady running it, who -- we're all just fantastic. She's doing a great job. She actually worked for us for a long time. She left and went to Santos. Now she's come back with a bit of [indiscernible] and doing -- just doing a great job. So I believe we're going to continue to see software sales accelerate. And that's including sales of interpretation software to oil companies, but also processing and imaging, which goes hand in glove with high-performance computing to small competitors. So doing projects that we don't really -- we don't want to do, they're a bit smaller than we would ideally like to do or they're in regions like the [indiscernible] or other places where we don't have a footprint. We don't do much work. And there's -- and the projects tend to be smaller and more difficult. We are working on a great strategy around interpretation, which is around getting into the larger oil and gas companies as well as getting the processing and imaging software out there into the broader marketplace. It is an exciting strategy, and I think watch this space. And of course, the third growth driver is the non-oil and gas stuff, which is DUG Nomad and the immersion cooling. I'll now start with the immersion cooling. I'd really encourage everybody to look at BAC's Cobalt new YouTube announcement. It is interesting, and it really gives you insight into how our immersion cooling goes hand in glove with their big cooling -- external cooling towers. And you can see how it simplifies the whole data center. It's really worth looking at. Of course, Nomad, we are really excited by. It is the mobile data centers. And I've said this before, but I'll say it again, there's a really big demand across all sorts of areas for Nomad. And some of it is being kicked off by -- when Russia went into Ukraine, the first thing they did was bomb out all the data centers. They did that to knock out communications. And that's one of the big focuses on Nomad and that's where it comes from. So, immersion cooling, whether it's in the Nomad form or inside your data center via BAC, is very compelling. It's really compelling, and we believe it will be the way data centers are built going forward. There are a couple of hurdles to get over, but it's extremely exciting. And of course, as I've said, to bring this all together, and to maximize our growth potential, we need to have a global footprint, and that's exactly what we do have now on the back of opening Abu Dhabi, but we've also pushed into India and into Brazil. I think all the paperwork signed for Brazil, isn't it, Dan? It's painful to get set up in Brazil. I think we're just waiting for that paperwork and agreements now to get registered and we're away. And in India, we're doing -- now both of these are different to the Middle East. In the Middle East, we actually started with the same sort of strategy, putting a BD front to find more work, see what we could find there. But in Middle East, we got swamped with opportunities and decided to put in a proper business unit. In India and Brazil, we don't intend to do that. It is a BD front that we're putting in. The Indian work will be done in Malaysia mostly and the Brazilian work in the Houston office. The order book is growing. And certainly, the signs are that it will continue to grow. We're excited by it, and we're seeing the pilot projects convert. We just continually cover the ground. We now see our competitors are now saying, yes, we're going to do MP-FWI. Yes, we have it, but they're years behind. It's a really good place to be. And now we're just getting our HPC in order to cut off -- maybe answer a question that's coming, no, we have no plans to buy compute right this minute. We have headroom in our compute, but we've got to get our data center ready to take because if Elastic takes off the way we believe, we are going to need more compute. But for the moment, we don't. The order book as it stands, what we're seeing, we have the headroom in the compute to do what work we've got. Yes, and we're seeing opportunities come from all the new markets. So, why invest in DUG? And to summarize it, this is the last slide. Game-changing technology in the MP-FWI, significant global pilot projects done to prove up that technology and great interest now from everybody, from the super majors down. Modern geoscience software package, broad, modern, Broad is important because it means if you go to other software, you've got to have different software applications and there your data is moving from one to another. With the DUG package, it's one package that goes from one thing to the next to the next. Tier 1 client base. As I've said many times, it's much simpler to tell you who we don't work for in the oil and gas space than who we do work for. Patent immersion cooling license, right? And that's -- we still believe we're going to see some revenue from that this financial year. We believe we're going to see Nomad sales this financial year. Order book is growing. It's back really strong and growing. We've got the cash in the bank. We don't need to raise money. We've got the capacity in the data center to put in more compute when significant growth occurs and we need that new compute. There's no plans to do it right now. We have a fantastic culture. We have -- everybody is excited. Everybody is looking forward. Everybody sees what's happening. It is an exciting time internally in DUG. And we're now really reaching into most corners in the world, certainly all the corners of the world where we want to be. We've got a lot of traveling down into Africa going now. We have no plans to start up permanent -- any permanent people or offices in Africa, but we were in attend -- for example, we attended a Nigerian conference in the last few weeks and came back with a whole string of opportunities that look great as well. So thanks, everybody. That's my quick run through the PowerPoint or our quick run through the PowerPoint. Thanks, Dan.

Thanks, Matt. Thanks, Dan. We might now move to questions. Just a reminder, if research analyst would like a question, please raise your hand. And we have a few questions online. Jack Daley at Shaw and Partners, if you could please ask your first question.

Congrats on the result. I think first one from me, good to see the, I guess, project awards that you called out in January. I guess when we think about the rest of the second half, is that -- obviously, 13.4% is quite considerable. How should we be thinking about the awards moving forward in the short-term and kind of building off the pipeline?

Yes. We're seeing it continue. We're still winning work and the -- thank you for the question, Jack, by the way. We're seeing the BD pipeline is strong. We're seeing opportunities and tenders coming through, and we're still winning projects. So it's a really exciting time.

I think you mentioned they're both kind of across Acoustic and Elastic. Is what regions are you seeing the most success at the moment?

London and KL -- not KL, London and Houston is where we're seeing the best uptake of this. And Middle East is just a bit more cautious. We've got several pilot projects in the Middle East and generating some massive excitement. And so I think you'll see Middle East joining them. But the projects that have been awarded so far have come out of Houston and London.

And I guess just on the -- I guess, if you won $13.4 million, probably implies about $4 million services revenue in January. Is that broadly correct? And I mean, is that just a function of January being a bit quieter and you'd expect that to step up?

I think January was $6 million revenue, wasn't it?

Yes, you've done your math correct, Jack. So for services work in January, that is -- it's the opening order book plus the wins minus the closing can tell you what we booked for services only in that period. And what we see is those awards through January ramping up really quickly. And so we expect that number to lift through the second half. It will lift through the second half as these projects come online.

But January is a very strong software month, Jack.

Okay.

So there is lots of renewals of software. So January -- software goes strong in January.

Okay. So it was probably 4 services and then 2 software, give or take? Yes, I think you had the strong software second half generally?

I don't remember the exact number for revenue in January, to be honest. It wasn't a weak month. It was a strong month.

Okay. And then I guess just on the HPC-as-a-Service, it looks like -- I mean obviously don't know what the first quarter number was, but it looks like it drops quite significantly. Are you able to give -- obviously, you kind of called out they went from -- they changed their way of kind of billing, it sounds like one of your biggest customers. Are you able to give a sense of what that looks like moving forward?

Well, it will build from here, I think. I was sitting in a meeting late last night with all the software and HPC global sales team. And we're looking at lots of opportunities. And I think the problem for that business, both in the non-oil and gas and the oil and gas, has been we had early wins and they come and go. They use more sometimes, they use less others, and that's what we're seeing come through the numbers. We just don't have the broad base of clients so that they all stack up to give a smooth result. There was nothing ominous about that result in oil and gas. That company just changed its policy and they didn't move anywhere else. So -- but the opportunities are there, and we're broadening that client base. So I certainly believe you will see it growing from here. And it is very important to the software sales as well.

And Alan Franklin, you've asked a few in the chat. I'm assuming you might want to ask those verbally. If you could go ahead with your questions, please.

Yes, happy to. Just a bit of detail on the Elastic workloads. I think this has been touched on a little bit already, but just in terms of the type of client that is looking to turn a pilot into real workloads. Is it really only for the scale of the super majors and/or is the reprocessing work or more sort of exploration on new assets?

It's -- what we're seeing so far is not the super majors, although we have one or 2 tenders in from the super majors now for Elastic MP-FWI. It is as per normal in our world, the early uptake as a technology is the big independent companies, especially the American ones. And that's what we've seen in a big one. And the projects are tending to be more on the production side than on the exploration side for the Elastic. The Acoustic is tending -- which is what we expected. We're seeing the Elastic on the production side a little more and the Acoustic on the exploration side. Some of the most excited clients though are in the Middle East as well as Houston and London. So the pilot projects are spread all over the world, but the early adopters is a company in -- the big independent in -- or 2 big independents in London and then a handful in -- a couple in -- out of Houston and more money in Houston.

Helpful. Just on the sort of cost base, relatively flat through the second half -- sorry, through the first half. Just how to think about any additional items to think through the second half? I assume that leased office comes on, but what other sort of lumpy cash growth -- sorry, cost of growth items should we think about?

So because of the way of lease accounting, the -- we signed that lease commencing 1 July. So the lease cost is actually coming through the books because although we had a rent-free period due to lease accounting, it's smoothed out as well as the staff there were mostly hired in July also. So most of those costs, there shouldn't be a big step up there. Obviously, hand over to Matt.

And some of the costs that we've incurred in the last 6 months is we've moved the shared services teams now completely up into Malaysia, and we're building them there. So that cost of redundancies and so forth are now behind us. We'll still be restructuring always, but the big chunks have gone. So unless you -- and we've got enough compute to do workload that we have, as I mentioned earlier. So unless we see accelerated growth, which is certainly what we're believing in, it's unlikely we'll be ordering more compute this financial year.

I think it's probably also worth highlighting, Allan, those Middle East staff through the period we're getting trained, they're all working productively now. We've been putting a lot of resources into the pilot projects. So on the staffing front as well, we feel like we have good capacity now globally to service an uptick in revenue.

Yes, helpful. And I might just drop in one query on the -- on BAC, just to what extent do you think they've actually formally live launched the product now? And just to the extent of how you think the first couple -- what sort of type of project the first couple of clients might be? Are these going to be more retrofit type solutions or more clean new builds that they might be pushing for?

I think the -- we don't have perfect insight. I think the early revenue will be probably more likely on the refit side just because those projects are quicker to get up, but they are targeting big new builds. That is what they're talking through, and they've presented the solution now to hyperscalers as well. We got confirmation. So they've -- it's a mix. I just think because of the time taken for a new build to come online, it's more likely that we'll see refit as the early sales and new build as the tail.

The follow-on.

The follow-on.

Okay. Ross Barrows, if you go ahead with your question, please.

Just some clarification around the Elastic offering. So if a new client explores Elastic and starts a trial that spans clearly all incremental. But with an existing client when they undertake an Elastic pilot, is that incremental spend on top of their current spend? Or is there some kind of cannibalization from their existing revenue they might be spending with you?

Yes, it's a good question, Ross. So I'm thinking of one particular client here. They had a project with us in the Middle East. And we're working the project, it's going well and then they just -- and it's a significant client that can be a humongous client, and they will just say we would love to see Elastic on a part of the project that we're working on. And so there's no -- we don't get paid for it, but we're very keen to show it. And we also don't want -- you don't -- in the early stages, you don't actually want production projects because you still just want to run a few projects through and sort out any issues that arise. So you want a little less pressure. So it's like that. It's -- most of the opportunities are like that or the stand-alone elastic little trials where you do get paid something, but you don't get paid very much. But they're set up differently. They're not set up as production projects, but now we've got the production projects coming through.

Okay. 2 other ones, they're both related, I guess. You mentioned now that you're planning to -- or preparing to fit out the balance of the fourth phase in Houston. You've been quite clear today that you don't expect that to be fitted out anytime soon, but things could change. But given the density and the capacity that you have put in recently is kind of orders of magnitude more capable than the existing installed base prior to that, for you to kind of chew through and work through that orders of magnitude, bigger compute with, I guess, the growth you've got to then expand the fourth phase by 20% and then deploy even more, must imply that you've got -- that would have to be contract wins of a very meaningful size to justify that final fit out?

Remember, Elastic uses a lot of compute and brings big dollars with it. And so you're just looking at a different world where a lot of compute is deployed, but it does bring big dollars with it. And so yes, we -- you can't win in a project unless -- you can win a project without the compute in place and then scramble like crazy to get the compute in place. You cannot win a project without infrastructure in place. So we believe strongly in where we're going, and we believe we will have accelerated -- sustained accelerated growth, we really do. And therefore, you've got to get that infrastructure in place because it cannot be done quickly. And you cannot -- you've got to put it in as part of your strategy. You can't put it in as part of winning projects because you're just way too late.

Just to finish that thought, to have enough work to go through the capacity you've recently deployed that then want to put more in would mean considerable visibility on demand?

Well, yes, we won't put more compute in until we really need it. That's been our mantra forever. But you've got to have the infrastructure in. And yes, we believe we're going to need it, and we believe we're going to grow. We are a growth company. We always grow. We do have flat spots from time to time, but we always grow. And on the back of the technology we have and what we're seeing in the market and the update from clients, we believe very strongly that we're going to have sustained accelerated growth, and we're going to need that capacity in the data center. Now -- but we don't need it right now. We've got enough headroom in the compute we have right now. But as I said, when you need the compute, it's too late to put in infrastructure. It's all over. Then you have to go to third-party, and we know how painful that is.

Yes, for sure. And just a really quick follow-up. You mentioned earlier about managing the extra heat that's coming off or being generated by the compute. Can you just quickly talk about, I guess, how you're mitigating that and any costs that might be associated?

Well, the AMD, the new AMD machines, you just have a lighter load in each tank, so you can't fill up each tank. Or you spread them around, you spread those new nodes through tanks of all the old nodes so that you average it down to what the tank can manage. In the infrastructure upgrade that we just finished, there is a kilowatt upgrade to each tank, but we were restricted because we already had a lot of electrical equipment and stuff in place for that. There was more electrical equipment put in, but we had the power boards at the end of the -- on the wall in place. We didn't have the distribution or the PDUs in place, we had the power boards. And we also had all the plumbing and subfloor plumbing in place. So we were a little restricted on what we could do with heat rejection and power there without throwing it all away and starting again. But they are -- I can't remember, there might be 30 kilowatts now, so they are upgraded, might be a bit more. In the new bit, it's not where we've got a little bit more of a blank canvas, it's not difficult. You put in bigger heat transfer units and you put in a different oil and you feed it more power. And you put in more chillers outside, more cooling towers outside. So that -- so the cost implication is not -- remember, the infrastructure costs are small relative to the compute costs. If we roll this out, this will be a really big deal revenue-wise because it's -- these are big things we're looking at.

Lindsay, if you could go ahead with your question.

Question, I mean, it feels like everyone's kind of trying to get -- trying to pinpoint this capacity question. I mean I know it's not this simple, but if we just have a look at maybe DUG historically, like I think the most you've probably ever done the service revenues is $30 million a half or something. Is there like a dollar figure or a dollar range that we could think of? Like your current capacity is sufficient to support $X worth of revenue, just so we can kind of frame it? Is there a simple answer you can give us?

No.

I figured that'd be your answer. Okay.

We get asked that question every time on DUG. Sorry, there's no simple formula. It depends on the projects and what you're doing and the clients and whether you -- the pilot projects, clearly, we're not getting -- we're getting less revenue than a production project or sometimes none. We're certainly aiming -- and on some of the win -- won projects, we're getting more. We're getting a premium. And then if the machines sit idle -- we run crypto not at the minute because everything is really getting busy now. But at the tail end of last year we've made $100,000 a month roughly from crypto. So it depends on where that revenue stream is coming from as to -- is it traditional processing imaging? Is it elastic? Is it -- what is it that's -- or is it some AI work? Is it crypto? What is it that's producing that revenue? They all have different revenue profiles.

Yes. No, that's -- I figured that would be the answer, but I thought I'd ask anyway. And then just -- I mean, the $13 million contract wins, like that's -- it's a pretty big number for a short space of time. I mean we've seen this in the past. Like if we go back to, say, last July, I think you had nearly a $20 million contract. Is the $13 million new wins, is that like a big contract and some smaller contracts? Is that like a good spread of medium-sized contracts? How should we think about the makeup of the $13 million?

There's some good spread, but -- there's some bigger projects in there and a good spread.

Okay. Brilliant. And then on Nomad, I mean, you raised some money. You've talked about building out the sales and marketing team. I think in the past, maybe you yourselves haven't been that clear on just the revenue model for the Nomad business. Like is there any more color you can give us there as to how that's progressing? Is it going to be a sale leases, those sort of things?

We're really disappointed that we don't have a sale yet, but it looks amazing. It does look really, really great. And the model is -- our ideal model is we lease the Nomad full of compute with all our software on top and manage remote -- and manage it for them. But then there's other leads, significant leads where they say, no, no, no, for security reasons or whatever reason, we want just to buy the Nomad itself and no compute and nothing else. And so in practice, we'll end up being a mix of those 2.

A question in relation to Elastic and the work that's been done over the first half. Did the focus on Elastic perhaps result in some of the slower pace of revenue being booked in the remainder of the quarter -- the second quarter?

Yes, it did. It didn't -- we didn't consciously do that. But you've just got some -- a very exciting new piece of tech and a lot of client focus on it. And a lot of questions being asked or do we do Elastic, what's the price point? Should we continue with our Acoustic? Is it time to switch from traditional processing and imaging? So there's a lot of noise being generated, a lot of focus. And so it did take away some of the revenue. And you've also got a lot of people working on those pilot projects and so forth. So a lot of support going into those pilot projects. So you would have slowed down a little bit the other projects. But it's all for the exactly, absolutely. Would you do it any differently if you did it over -- no. It's all for the right reasons.

Okay. Jack, returning to you if you have any further questions.

You kind of flagged the multi-client, I guess, expansion into that kind of market. Can we get a sense of, I guess, sales capacity required, what the investment is from your side? I guess, how does that work potentially cross-selling or leverage your existing relationships and when kind of revenue contribution might look like for that?

Yes. Thanks, Jack. So first thing, we've been in doing multi-client for a long time now, but our strategy has been to take a little piece of other people's projects. So we work for 12 multi-client companies, and we will take risk on the processing in order to get upside. So that's been our take on it. And there's an obvious flaw with that model that you have to be very careful with is if they've got a well underwritten, really exciting project, then they won't give us a share of it. They'll just pay proprietary processing and imaging fees. If they've got a project's a little more risky, then they will probably give us a share of it. So we decided to do it ourselves as well as work for them, and that's working really well. The key here, though, is that when you do a multi-client project, you don't just go out grab some data, process it with no clients and then go and try to sell it. You build up a project, you generate a project, you document it, you say what you're going to do, have all your plans in place. And then your sales guys go out and show it to the oil companies and try to get what's called pre-funders. So we're looking at several projects. One in particular I've got on my mind is we need 2 pre-funders before we kick off the project. And those 2 pre-funders will cover our costs. So once -- so we don't have a risk here. We're not taking risk. We're not going to be making big investments into this business. We're going to be willing and dealing to get really lovely upside, but without putting capital at risk.

Lindsay, I think you have one more question?

No, I don't think so. I put my hand up because it's a mistake.

No problems. Thank you. I think we have addressed all of the questions unless there are any that I have missed. Matt, would you like to make any closing remarks before we leave? And before I hand over to you, I would just like to remind everybody that Matt and Dan are in Melbourne and Sydney next week. Melbourne on the Tuesday, Sydney, Wednesday, Thursday, Friday. If you would like to arrange a one-on-one meeting, please get in touch, and we will make that happen. Matt, any closing remarks?

Yes. Look, I'd like to thank all our investors who have stayed with us through quite a trial over the last 6 months. We are working very hard. We have everything we need for sustained growth, and I hope you hang in there with us because I think it's exciting times coming. But thank you very much for hanging in there with us over the last 6 months. And hopefully, I look forward to meeting a lot of you next week, face-to-face.