DUG Technology Ltd (DUG) Earnings Call Transcript & Summary

Who will take us through the presentation, which was lodged with ASX earlier this morning. We expect this to run for about 30 minutes, after which we'll have Q&A. [Operator Instructions] Handing over to you, Matt, to take us through the presentation.

Thanks, Steve. It's great to be with everybody again at the end of another financial year. I suspect everybody is getting used to who we are. So we're the data people. We're the numerical analysts. We do big algorithms, and we have big high-performance computers. We're about 21 years old now. And our offices are in Perth, London, Houston, KL and Abu Dhabi. We are growing employees, but slowly as we're doing it very carefully and looking at what's going on. We have more than 45 petaflops of compute. And you'll see in my note, I talked about we've quadrupled our compute, and we've done that via adding new compute that everyone is aware of, but we've also taken the RAM in over 9,000 machines we have from 128 giga RAM to 384 and that's given them a lot greater throughput. When you're comparing us with our competitors, you'll see our competitors often quote single precision petaflops, so you can double our number to compare with their number. Thanks, Steve. So this is our journey, and everyone has probably seen it before. The nice bit is on the end there, where we now have an Abu Dhabi office, which is where the fit-out is about to kick off. And I think we've got 16 employees over there now who are about to get training and learn our best practice and so forth. Thanks, Steve. So services is where we make most of our money. That is services [are in] oil and gas, they're in offshore wind farms and a few other bits and bobs where people do seismic acquire seismic data and we process it for them. That's where the multi-parameter FWI Imaging is. Yes. So that's where we make most of our money. Software is a great business. It's our favorite business. It's $11 million in revenue this year. And that's the software that we use, but it also looks after the interpretation piece. So it's the piece that is done in the chain after we process the data, it's interpreted generally by oil companies themselves, and they use our software to do that. And then high-performance computing is that we just let people come on and use our computing. They can come on and they can use our computers using their own software. They can come on, use our computers using our software. They can come on and do it in whichever way they like. And we can support them and help them with their codes there as well. Thanks, Steve. So oil and gas, I've mentioned, is where we make most of our revenue. I think it's probably the vast -- well over 90% of our revenue comes from oil and gas. Enterprise, we have a lot of clients. At the moment, they just don't spend that much money with us. But all the bioinformatics institutes from [ CSIRO ] to climate modeling to lots of stuff is going on in that place -- in that space. And then national security space, we don't have any revenue yet, but we're still chasing that down. We're actually in a big conference in Adelaide this week. Thanks, Steve. So this is the world according to DUG. The orange-colored countries are where we have worked -- we've done projects before. So you can see we've worked most places in the world. There's a dotted line running across the world there. That's our fiber network. We don't own it. We just lease it. And yes. And then we've got our machines around the world. There's 3 big data centers now around the world. But yes, it's great to see you Abu Dhabi on the map there now. Thanks, Steve. And look, we want to be the biggest and best seismic processing and imaging company on earth, and we're working towards that goal. There's a lot of market for us to gain to get there. We've got our sustainable supercomputing, and we grow that, but we grow it judiciously as we go along. And we want to really make a mark in high-performance computing as a service as well. So -- and continue to do our R&D. The BAC announcement showed you what value can come from our R&D efforts. Thanks, Steve. So getting on to the business of the day. So service wins for the year of $67.4 million, which is a really, really big step up from what we had in the year previous. Revenue of $65.5 million, so up 29%. Underlying EBITDA of $23.2 million, which is terrific. And of course, the difference between our actual EBITDA and underlying EBITDA is third-party compute, which because of circumstances of how we won work and when -- how we won work in a really big focused time last year, we got caught with not enough compute. And that meant we had to use third-party compute, and that's the difference between the EBITDA and the underlying EBITDA. The Services order book at the moment is $36.5 million. And look, that's a really strong order book for us. What we have is a really large pipeline at the moment. So we're waiting to see what happens to that pipeline. And people are just taking a little longer this year compared with last year to award work. But it's -- but the pipeline looks terrific. And so the pipeline, just to remind everybody, is where we put a proposal -- well, there's 2 parts of the pipeline. This work is where we've actually signed contracts. So this work will be done -- it's underway now. The pipeline is then proposals that we've written and delivered, and we can -- we risk those as the BD guys will risk them as to whether they think we'll win or lose those particular jobs, and we keep that in the first part of the pipeline. And then the early part of the pipeline is just where we get leads and where we do market surveys and we put them into the early part, and they're risked very heavily. That pipeline looks -- we've never had a bigger pipeline that it looks quite amazing. Thanks, Steve. So over to Ajesh.

Thanks, Matt. So the key takeout on this slide is that the strong growth in underlying earnings, EBITDA, excluding $6.5 million cost of third-party compute costs is up 54% on prior year. Total revenue growth was underpinned by services revenue growth of $14.4 million or 36% to a record high of $54.7 million. DUG won more service contracts than anticipated, which increased our compute requirements, and we chose to meet this extra demand by using third-party compute whilst our compute was delivered and commissioned. The cost of this was incurred from November '23 to June '24 of $6.6 million, and it doesn't continue into July '24. Excluding the third-party compute costs, other operating costs grew by $4.2 million or circa 40%, driven by IT and data center facility costs, up $1.6 million and growth in sales and marketing up $0.5 million or 42%. Thanks, Steve. Three points on this. PPE, property, plant and equipment has increased by $26 or so million, reflecting the new computing auxiliary hardware, less depreciation. DUG has a net debt of $14.5 million being the asset finance liabilities of $23.9 million, less $9.4 million in cash. Our strong cash generation will enable us to prioritize repayment of the debt facilities in the following financial year. Thanks, Steve. We generated strong operating cash flow of $12.1 million. This would have been $16.5 million without the third-party compute costs paid for in the year. About $7 million of PPE, property, plant and equipment was funded from our own cash flows that are not financed. And just a reminder that the capital grant that was received has been recorded under contract liabilities on the balance sheet and has not been taken to the P&L. Thanks, Matt.

So looking at breaking our business down into the 3 areas, Services. So look, Services is really strong. You can read the numbers there, record high. We expect it to grow with the new office in the UAE. We've got new technology out, which is the Elastic MP-FWI and the first results of that look very encouraging. We'll be showing them at the big conference in Houston. I'm over in Houston, everybody for our biggest conference of the year is next week here in Houston. And so we'll be showing our first results from that MP-FWI -- Elastic FWI and it's very exciting and look terrific. So we expect to be able to drive growth in services through the new technology as well as the new office in the Middle East. Thanks, Steve. Software, we ended up growing by 11%. And look, this is the business that we really love. Everyone loves the Software business. And we will -- we've got a new global BD manager for this who's actually rejoined us. She worked for us a few years ago before she went and worked for [ Santos ] for the -- interleaving years. She come back, fantastic, fantastic person, and we're really looking to grow this business in the coming year. Thanks, Steve. HPC as a Service is a bit disappointing, but the highlight here is not in the last financial year. It's in the winning of a $500,000 deal for this year in July for this financial year. So that's a really great news for a business line that was -- that we weren't particularly pleased with. We have reshaped it a little bit. And yes, we're still confident that we're going to grow. It's just about timing on this business. Thanks, Steve. Look, we've got -- sorry about the lights. We've got really, really strong relationships with our clients and the relationships grow over time. They get more confident in us as we do projects. And as they get more confident in us, we get larger projects and we get more complex projects. So we're growing our revenue from each client. We virtually work for every oil company on earth. So it's now about doing more work for them, doing better work because we're still 6% or 7% of the market is all we have. So there's an awful lot of work we're not seeing just yet, but we're seeing more and more of it as we go along. And we build our relationships. Client A there, we had a meeting with just last week, and they're just extremely happy with us. That's a client from the Middle East, but that was one of our existing clients from the Middle East where the work is done in Houston. Thanks, Steve. So again, I think everybody realizes that Middle East is the player in the oil and gas space. They are ramping up big time, big time, massive surveys, bigger than I could have ever dreamed of, lots of activity, very exciting. We're sort of just there, just in time, if you like, probably a bit late, but really great. And we did win a project with Saudi Aramco, which we're really pleased about. And the pipeline in the Middle East is really large. Jay Van Berlo is there for us, and he's just doing a terrific job. So we're really excited by that office. And the caliber of people we've got there is really, really great. We've managed to cherry pick from other offices in the Middle East, and it looks really good. Thanks, Steve. Yes. So this -- we're really excited by this deal. This -- we've been working on this deal for a long, long time and seriously working on it for 18 months, but it goes back another 6, 8 months before that again. And basically, BAC -- well, first of all, BAC sell the majority of cooling towers for whatever purpose the cooling towers are used in the world. So whether it's a building, building will have cooling towers or whether it's a data center or whatever the purpose is, BAC sell more than 50% of them. They have 500 channel partners globally. They have, I think, 8 manufacturing plants. It probably says that they're somewhere. I mean, this is a huge private company. And so we -- this is not just another small company selling immersion cooling. This is a really big established company. They have a very large data center segment in their business. They already have deep relationships with all the hyperscalers. So these were the guys we wanted to do the deal with, and they're the guys we went after. So they have an exclusive deal. We get a 5% revenue of net sales. We can still use the IP ourselves. They're going to take over doing all the R&D on the tanks because the newer computers are using more power and therefore, they produce more heat and so the tank capacity in terms of kilowatts needs to go up. They're going to do all the R&D. They've already started. They started some time ago. They're going to do all the defense and pay for all the defense if there's any patent defense required. So they're taking over this completely, and we just sit back and get our 5%. What we need to do is just to show people through our data center, especially the one down here in Houston to show them what immersion cooling looks like at the scale, talk about our experience with it and share numbers that we already have. So it's very not onerous on us at all this deal is a beautiful deal. Thanks, Steve. So DUG Nomad is a business unit that we've just started up. We've actually got one of these Nomads in Adelaide this week at the Defense show. We've got another Nomad that's on its way to America to a big client here. And so we're just looking at now commercializing this. It's a really cool piece of kit. So whereas with DUG Cool -- excuse me, for a minute. With DUG Cool, we chose to -- we just needed to be much bigger than what we were to do justice to DUG Cool. Nomad, I think, is a really good fit for us. And so we've started a business unit. We brought in a new guy who's a CEO of another company. We brought him in to run that new business unit. And yes, we're really excited with where Nomad can go. Thanks, Steve. So look, we've got really outstanding organic growth. We believe we can continue our momentum. If you look at what's going to drive our growth, we've got the new office in the Middle East. I think that office will be our biggest office from what we're seeing in the pipeline. We've got a strong -- we've got a very strong pipeline. We've got a strong order book, but we've got a very, very strong pipeline. And we've got really exciting new technology in the Elastic FWI and the Elastic MP-FWI, and the first results look really exciting. And on top of the Middle East, we're also receiving lots of tender opportunities out of India now. We've done a lot of work in India. We're now -- we're in about the same stage in Brazil. So we're not just expanding into the Middle East. We're also expanding into India and into Brazil. And just offices have had just terrific years like the Malaysia office, [ Gunnar ], who runs the Malaysian office has done a great job, and that office has grown at over 100% this last year and is continuing to grow. And now he's -- the Malaysian office is now the office reaching out into India. So we're exciting by where we're going in this financial year. Thanks, Steve. I think that's it for our presentation today.

It is. What we might now do is move to Q&A. A number of people have put up their hands. What I propose we do is run through. [Operator Instructions] We have a little bit of time because Matt has kept things nice and tight. Thank you, Matt and Ajesh. Perhaps we can start, Julian, if you could go ahead, please and ask your question.

Firstly, just on the order book map, it kind of dipped down a little bit in that sort of fourth quarter and then the win rate was a bit lower. So you said that people are taking a bit longer to award contracts. When do you think you'll start to see this step up? Would it be kind of this half or towards the back end of the financial year?

No. We expect -- well, everyone's got to come back from holidays. August is a big holiday month in the Northern Hemisphere. And so we need everyone to come back from holidays, and then we expect to see action start. So we expect to see work to be coming in pretty promptly.

Of course. And then on the third party...

It's not an unusual thing, Julian. Sorry, Julian. This is not unusual. It's sort of like awards, wax and wanes it's -- so this is not unusual. We did lose more than we would normally lose over the last 6 months, but that's because we put our price up because we had more work than we could handle. So it's all very predictable at the moment.

Right. And the margin on the MP-FWI, is that kind of sort of landing where you thought, higher than the traditional sort of form?

It's mixed. It depends on the marketplace where we're selling it into. We get more money for it in America at the moment than we do out of the London office, for example, but we're working on getting that price up everywhere. And we're also -- that code is under a heck of a lot of work. We've got a big team working on that code. So it's getting more efficient as well. We're about to roll out a 10% efficiency gain and there's more coming. So we'll be -- I think that margin will increase.

And just finally, so with the third-party compute, the cost of $6.6 million, what revenue would be attributed to that $6.6 million that you had to sort of outsource?

Yes, I can't answer that question, Julian. I don't know. It's not how we sort of think of it, I'm afraid. And I don't know the answer to the projects that we run on it. We did just run MP-FWI on the third-party compute. So I don't know the answer.

Lindsay from Ords, if you could go ahead and ask your question, please.

Hope you can hear me?

Yes.

Yes, Matt, a couple of questions for me. Just on MP-FWI again. I think in the past, you've called out that MP-FWI was something like 1/3 of the new wins in certain periods. Just wondering kind of where that is today and maybe how we should be thinking about MP-FWI as a percentage of your wins, say, over the next 12 months?

It's been 1/3 of our revenue over the last 12 months, which was really good to see. I would expect it to grow up from there, to be honest. And I think we'll see more wins coming. The Elastic is more expensive again. And so that's another reason why I think you'll see it dominating wins and revenue as we go forward.

Okay. Brilliant. And just timing maybe on when we start to see commercialization on the Elastic front?

Next week at the big conference, we'll be showing these results that look terrific, and we've got a lot of clients lining up to look at the Elastic. So I think that will -- that shouldn't take too long. Remember, it does take -- when I say it won't take long, just from the -- start talking about a project to write a proposal for a project to get the project awarded is you're talking -- in general, it varies a lot. But you're talking 3 or 4 months. So we're talking pretty promptly. But when I say promptly, we're talking 3 or 4 months probably.

Okay. That's helpful. And then just a final question, maybe more structural. I mean I think you've always kind of talked to the business delivering something like 30% EBITDA margins longer term. The second half, you've delivered something closer to 40%, and that includes scale up in the Middle East and these sorts of things. If I look at Viridian, their geo business does north of 50% margins. Just trying to understand, is 30% still the right margin to work with longer term? Or is there something that's fundamentally changed where maybe you can start to deliver better profitability in the business?

Viridian is a business that charges a lot more for their services. So they're like -- if you think of you don't get in trouble for buying IBM, at the moment, they're the IBM in the industry, if you like. But that's being broken down every day. I talk to a client and they're getting disappointed results. And so there -- but that's where they've been sitting, although it is breaking down now. So their margins will be higher, but they are coming down. They certainly will be coming down in my belief. So do I think we can increase our EBITDA margin? Yes, I do, but I would be I'd be a little cautious to say what it could go to, but I do think we can increase that, yes.

Ross Barrows from Wilsons, if you could go ahead with your question, please.

Can you hear me, sir?

We can hear you.

Yes, a couple from me. Matt, you mentioned that you're attending a bigger conference for the year or the biggest of the year next week in Houston, and you and maybe some of your colleagues have attended others over the past few months. I was just wondering if you could share any color from those events so far and maybe making any observations, I guess, around your product versus the competitive landscape?

The Elastic has caught people's attention as well as our MP-FWI, but they're very, very, very different offerings. And people are excited. The whole industry is up and about, and people are very excited by where FWI is going. They're very excited by the Elastic. They're very excited by MP-FWI. And they're very excited by the 2 coming together, which is what we've got now. Of course, we're the only people with that. And as a guy from Shell said at the big European conference maybe 2 months ago now, he said, you've got the holy grail. And where you feed in the field data and you get out everything you need all the way through to the end of inversion, which we would typically do as a separate step traditionally. So everyone is really excited. It's a really fun time to be in as a technology person, it's a really fun time to be in the oil and gas industry. It's up and about lots happening and lots of excitement.

Great. And just a follow-up. You mentioned the word pipeline, and you haven't quantified that and probably not expecting you to now. But just in terms of order of magnitude of how that looks today versus, I guess, how it's looked in the past. Can you give us any kind of guideline around that, even not saying what the number is, but just the size of it relative to how it's been in the past?

Yes. It's really big. It's way bigger than what we've had in the past, way, way bigger. But it's also based on a lot of Middle East work. A lot of the pipeline is Middle East. The pipeline out of the Middle East is very -- looks normal, looks like what it has been. And it's good. It's very solid, but it looks like what it is. The Middle East part of it is huge. And the issue we have with the Middle East part, of course, is that we've won one job. We've got some traditional clients in the Middle East, but not a lot, but our biggest client is a traditional client in the Middle East. But they're sort of like a company which has a Western way to get into it. There's Westerners in key positions in our clients there that we know, whereas the big ADNOC and the Aramco, they're like more seriously Middle Eastern companies. And we don't understand how and when they award jobs. We have no great feeling for how our pricing sits in the market yet, although we won one job, we haven't -- we need to win and we need to lose and we need to start to understand what we're going to get out of that. We haven't lost any projects in the Middle East. We've won one since we've been putting out tenders. Yes, we lost one actually. We did lose one. So we won one, lost one so far. And we just need to keep doing that to start to understand what -- how it's all going to fall out.

And just super quick, just to characterize the Middle East clients, the kind of orders of magnitude bigger than, I guess, the existing clients that you've been working with. Is that a fair summary?

I can't actually -- I don't know that. I mean they're the biggest companies on earth, I think, in Aramco and ADNOC, they're huge. How much bigger than an Exxon or a Shell or BP, they are, I don't actually know. What I do know, Ross -- sorry, Steve, what I do know, Ross, is that the seismic surveys are in order of magnitude more expensive than the other surveys we're seeing elsewhere in the world, which is the most important thing for us.

Jack Daley from Shaw and Partners, if you can go ahead with your question.

You can hear me right?

We can hear you.

Perfect. And congrats again on a great result. Maybe I noticed a slide in the deck just around the top 10 customers and the contribution there. It seems like it's stepping up there in terms of the revenue that you're taking from your top customers. Could you maybe just talk about that a little bit in terms of, I guess, getting your foot in the door and then winning more work in bringing up the tech there? And how do you see that trending over time?

Yes. So this is Steve's favorite story actually. So there you go, Steve, you got the question you would like. Yes. So when you start working -- as I said, we sort of work for every client on earth. We don't necessarily do much work for every client on earth. And when you start working for a client, you will do a simple small project in a non-core area is typically how it would work. And that's after doing a whole bunch of tests and trials to prove your technology is okay. And then as you prove yourself and you do more projects and they get more confidence in you, then you'll do a bigger project from a non-core area. And then you'll do a small project from a core area and you'll do more sophisticated work and you'll do more production work and you work your way up the chain as you go along. And we are working our way up the chain, and you're at different positions in that chain with different clients. For small companies, you can go to the top pretty quickly. It doesn't take long. For the large guys, it takes years and years, and they just put you through the ringer. And -- but we're well on the way up now. And yes, but that's how the relationship works between us and clients.

Great. And maybe just on -- I guess you mentioned the other contract in the Middle East, you didn't win. Can you give us just a sense of if you know who they went with and why? And anything you need to do different there?

So we always follow up a loss and understand why we lost and who it went to. And usually, they don't tell you who it went to, but you hear it on the street. But we haven't had feedback yet on that one. That's a fairly recent loss, and we haven't had feedback of why we lost or where it went to.

Okay. Maybe just -- sorry.

Yes, sorry, Jack. Next question.

And just last one for me. Just you mentioned Elastic is a fair bit more than MP-FWI. Is that -- when you go to market, how does your offering stack up versus the other guys in the market? Because obviously, MP-FWI, no one else really has it. But when you go live with Elastic, how is it going to kind of compare to the other guys that are already in market with that solution?

So the first thing, Jack, is that Elastic can be single-parameter Elastic or multi-parameter Elastic. So we brought out single-parameter elastic maybe 6, 7 weeks, 8 weeks ago. And we've brought out multi-parameter elastic maybe 2 weeks ago. So that's the first thing. It's -- they're quite separate things. And our -- it's going to be really interesting to see how we stack up because we haven't had it for very long, we don't know quite how it stacks up against our competitors. But what we're hearing, I think it will stack up really well. But what we're seeing in our test results is that you get a step-up in the quality of the result going from the multi-parameter acoustic to the single parameter elastic and then you get another step up of about the same amount going from the single-parameter elastic to the multi-parameter elastic. So if that holds true, then we're in really good shape with the multi-parameter elastic. In other words, if, say, Viridian's single parameter elastic sits at about the same quality as our single-parameter elastic, then our multi-parameter is going to be the best one in the market. But this is all to be tested, and it will fall out and we will learn where we're at.

Allan, if you could go ahead with your question, please.

Matt, have you got?

Yes, Allan.

Yes. I guess, late night, we are over in Houston, but just a couple of ones just to step out a couple of the earlier questions, please, maybe for Ajesh. But just to understand the OpEx piece, I know it sort of around about the first half, you had sort of 250 employees. You now have sort of 270. Just trying to understand whether that is in any of the numbers in FY '24? Or does that more build into the first half?

Thanks, Allan. Yes. So look, that step-up in employee costs would be the second half, and there has been incremental costs on the back of the 600 computers coming on board at the start of the half with regards to electricity and other data center costs.

Yes. And then just a quick follow-on. With the Middle East office and the team there, to what extent have they hit the ground already? Or is that cost not in the numbers as yet?

No, there's no cost for people as yet in the FY '24 numbers. We've just got Jay Van Berlo on the ground who has been there for this half.

Perfect.

Those guys only joined in August, Allan.

Helpful. And then perhaps just a query or just a follow-up in terms of the work in hand and/or how to think about sort of timeliness of execution for clients. Have you seen -- maybe if you can just talk to the delivery timelines if that has come in materially and/or in this current world, do you need to be delivering throughput to clients in a lot quicker manner, hence, that sort of order book and the work in hand turns a lot quicker than perhaps it may would have a period ago.

The MP-FWI work turns over quicker. The traditional projects take the same amount of time really as we're all trying to do them quicker and they probably they've sped up in the years gone by. But at the moment, to be honest, I don't -- I think they're all tracking about the same as what they always have. So it depends on what sort of work it is.

Perfect. And then just one last quick one, just in terms of the extent to which FWI is sort of cannibalizing some of the core work. I mean if you just strip it halfway back, there has been a little bit of a decline in the core work, but it doesn't look like it's gone materially back. A lot of the FWI seems to be incremental services growth coming through. Is that a fair observation?

Yes. I think -- I don't think that we are cannibalizing our traditional business just yet. I don't -- I mean, that's what we want to do, of course. Then with our employee base, we can produce a heck of a lot more revenue than what we are now. But yes, I don't look at it that way just yet. And of course, the Middle East work is all land, which is a lot more difficult in the FWI front. So yes, I'm not getting the feeling that we're cannibalizing stuff just yet.

Caleb Weng, if you can go ahead with your question, please.

Congrats, Matt and Ajesh, on a good financial year. So I think Viridian commented last week -- or last month basically saying that the Middle East companies, they've kind of spent the past few years gathering seismic data and now they're kind of going to start processing and imaging that data. And you also mentioned earlier in the call that you feel like you kind of timed your entry into the Middle East perfectly. So I guess, can we interpret that as sort of a lot of these Middle East oil companies, they kind of -- the focus for the exploration budget a few years ago was more towards data acquisition. And now for the next few years, it's going to be more processing and imaging or...

I think that they've been doing a lot of processing and imaging over the last few years. I think that they are going into a new phase of processing and imaging, but they're also still acquiring massive surveys, the biggest surveys that have ever been acquired on earth. So they're very -- the things are going on in parallel, but it is really exciting what's going on over there. I think what's happened a little bit in the Middle East over the last few years is that there's been a lot of cheaper processing done. And I think now they're looking at really doing their processing really well. And so we -- that's probably more the way I would look at it. It's a new era of processing. And just while we're on Viridian, they're the guys that say that because I often get asked, they've got 500 petaflops, you've got 45 or 50 petaflops. Well, Viridian use single precision. So if you want to compare our numbers to there, they've got 250 versus our 50 is the comparison to make computer-wise.

Yes, for sure. And you guys are probably a lot more competitive in the other higher-end processing stuff.

Yes, we are.

And I think you mentioned earlier in the call that you're kind of cautious about adding headcount. Is that because, I guess, you see kind of MP-FWI freeing resources as that get utilized more? Are you just being cautious and not trying to, I guess, add headcount before you win some more contracts?

No. We are actively hiring. We're cautious that we're getting the right quality of people and that we're really being careful and judicious with who we're hiring. And we certainly don't want to balloon the company out people-wise. But Houston, for example, we're hiring and we need more people in Houston. And yes, we do see FWI taking over, and we do see -- meaning we don't need as many people as we would to grow the revenue through traditional means.

Okay. Thanks, Caleb. We have a question that's coming via the Q&A chat channel. It's one in relation to the new compute that's recently been installed, Matt. And the question was how much additional work or how much additional revenue, however you measure it, does the additional compute enable you to handle? Or asked another way, how long before you think you might need to add more compute?

So the -- currently, the -- we have a bit of spare headroom on the compute, it looks like, which is really good. So we've got to where we want to be, which is a bit of spare headroom. And so we have enough compute for what we're foreseeing at the moment. If we win a bunch of this work in the Middle East or we win a whole bunch of Elastic MP-FWI work, then we'll need to add more compute. But that's a good scenario, right? That's -- but if we don't -- if we get the workload that we're looking at, at the moment and we keep growing that workload steadily and it looks like what it is, then we'll be fine for compute this financial year. So it just depends on how that order book goes and what the order book is comprised of as well. And in terms of revenue from the compute, again, it depends on what sort of -- what we're running on the compute, whether they're running and the balance between the cost of a project for us, generally, the main cost is people and compute. And if it's traditional processing, it's very people hungry and more lighter on compute. And if it's MP-FWI, it's very people light and very compute heavy. So it depends on -- and the revenue is not that difference between the 2. So the -- it depends on what the makeup of the work is on the machines.

Okay. One more question has come in, and it might be worth just touching base on this. Revenue expectations around DUG Cool and DUG Nomad, are they're still relatively early stage, but any commentary you can add to that would be helpful.

Yes. Look, I think I gave an answer the other day just to give you some sort of feel. I think the question was compared with services today, I'm not sure they meant to ask the question this way, but this is the way it came out. They said, compared with the services profit today, will DUG Cool make more money in 5 years' time? And the answer I gave was, I believe in 5 years' time, yes, DUG Cool will make more money than our profit today easily. But will it make more money than services in 5 years' time? I certainly don't think so, no. So I believe that both are going to grow, but services has a really large upside, we believe, and for a long time to come. But the thing about DUG Cool, of course, is it's just revenue goes straight to the bottom line. It has no cost associated with it. It's a beautiful thing.

And any comments on DUG Nomad?

DUG Nomad, again, very early days. We've got some great leads it could be significant. But again, it's very early. We haven't sold any yet. So we have to wait and see.

Okay. We might return to analysts on the panel. Jack Daley, if you could go ahead and ask your question, please.

Just a quick one. Just on the cash conversion, just looked a little bit lower than what it's been traditionally. Are there any comments there, Ajesh?

Yes. And that's particularly evident in the last quarter. The debtor book has grown. We've had billings and the cash doesn't come in yet at the reporting date. And just other working capital growth requirements.

Okay. Noted. And is that not really going to be an issue moving forward or...

We don't expect so we expect to generate strong cash flow and cash conversion in next year.

Okay. Lindsay, if you can go ahead and ask your question, please?

Matt, just another question on the Middle East. I suppose my question is, you sound very confident on the Middle East opportunity, but you've always sounded very confident in the Middle East opportunity. What feels like it's changed is that as at June 30, you had one employee in the region and now you have many multiples of that, I'm led to believe. Just trying to understand kind of what's changed. It feels like you've gone from confident and not allocating resources to confident and allocating resources. So that feels like a bit of a step change. Yes, just trying to understand what has changed to make you more confident on the Middle East opportunity in, let's say, the last 3 months or so.

So we put Jay into the Middle East as the BD guy in January. And our intention was to have a BD team and add a second person in there to work for Jay and to win work and to put that work out into our London and Houston offices and Malaysian offices as well. That was the initial plan. But as soon as Jay hit the ground in January, within a month, 2 months, what we were looking at was significant, like extremely significant amounts of work. And the way we were welcomed into the Middle East was wonderful, right? And so I sat back and thought, well, we're underdoing this. We're undercapitalizing this. This opportunity, I've never seen before. I've never seen a welcome like this. I've never seen anything like the amount of work they have and the amount of work they're going to have here. We need to do -- we need to change our plan, and we need to invest money on this opportunity to really capitalize on it. And so that plan was fairly early in the year. We just changed our mind and decided to put in an office and geophysicist. And then, of course, you start with, well, where will we put the office? And how many geophysicist to give you -- you've got to have bench strength. You can't go in with 5 geos and think that people are going to take it seriously. So what is the number of geos that you need and where you're going to find these geos? And then you start sending in teams of -- we sent a team of 3 HR guys and Jay into Egypt. Frankly that would be 4, 5 months ago now, and they went and interviewed like 85 geophysicist in Egypt, and then we went around the place. And it just all takes time. So it might look like a couple of months ago, we started doing an office in Abu Dhabi, but no, it's a long time in the making. And then you've got to get job offers to those guys and then you've got to get work visas and then you've got to -- there's an enormous amount of work has gone into getting that office to where it's at. And then, of course, we have to we go there and we stop around Abu Dhabi and look at all the different tax regimes and where we're going to put the office. And so we're in a 0 tax regime region in Abu Dhabi, and that comes with a whole bunch of overheads. So you've got to deal with those overheads and so forth. So there's a lot of work gone into that new Abu Dhabi office, and it's taken 4, 5 months of a solid team working on that to bring it to where it is today. In terms of numbers, we're sitting, I believe, at 16 geos now in that office.

Okay. Brilliant. No, that's really helpful color. And then just on Nomad, maybe feels a little bit preemptive, but some of the conversations you're having, I mean, are they like single unit type conversations? Or when and if you actually do sign a DUG Nomad deal, should we be expecting tens of units, hundreds of units to be sold? Just how do we think about the quantum there? That would be helpful.

Yes, it would be really helpful. Look, these are just rough leads. Until we start selling things and we start understanding the market better, a bit like the Middle East, then we'll know a lot more. We've got a lot of leads that are 1 to 2 to 3 Nomads. And we have one lead in the States here, which is about 60, 61 Nomads. And then there's a lot of crypto miners want them. And so it can be really big, but we need to see that what we're offering the market wants and they want to buy it. We think it's really cool, and it is unique. So -- but do people want it, will they pay for it? And that's where we're at, at the moment. And the people that we're sending -- it's a big company we're sending the Nomad into the states is I like -- it's like a channel partner, if you like, and they're really excited by it. So I think the feedback is really positive, but I'm not going to get too carried away until we start to sell them.

A question is coming online, an update on Geraldton?

Yes. So Geraldton, look, we -- you can think of Geraldton and that you've seen our -- we're disappointed with our HPC as a Service result in Australia. And clearly, we're disappointed by that. Everybody would be disappointed by that, except for the new job we've just won with the Astrophysicist. And so that goes hand-in-hand with -- if you're not building that business the way we expected it to build, we still believe it will go up. Do you need a really big data campus like we were planning in Geraldton unless you're building a really big HPC as a Service business. So those 2 go hand in hand. And so because of the momentum wasn't in that business unit, we said, well, it doesn't make sense to go ahead with Geraldton unless we have an anchor tenant. And so we went back to the state government and said we really need an anchor tenant. And so that's what we're doing now. So we've paused it all, if you like. We've still got the option to lease on the land there, and we've paused it while we're looking for an anchor tenant. And the state government is extremely keen for that project to go ahead. So they're working with us, giving us time on that grant, no problem at all. And they're actually looking within themselves for whether they can provide an anchor tenant. So I can't tell you any more than that, whether it's going to get up or not, I'm not sure. But we -- there's a number of things happening that give it pause for thought that it may well get up.

Okay. We might go to our final question, Ross Barrows from Wilsons, if you could go ahead with your question.

Just a clarification or maybe an extension of a comment you made earlier, Matt, in terms of DUG winning more work and you've got some headroom to execute on that as it stands. But just in terms of adding more compute, could you answer the question in 2 ways, one in terms of physical space and then one in terms of compute capacity. So you might have 50% or 100% more space, but given the density that you might be able to put in there, it might be 200% more compute. So could you maybe just give us some flavor around that?

That's a great question, Ross, and one I'm delighted to answer because we've been doing a great deal of work on this. So although we are not buying compute and have no plans to buy compute until we win a lot more work, we have -- don't want to get caught with our pants down again. And so we've been doing an awful lot of work. So -- the data center in Houston here called Skybox is almost 50% full. So we've got 50% more space to grow into. But of course, new computers a lot more dense than old compute. And so although we have about 50 petaflops at the moment or 45 -- between 45 and 50 petaflops at the moment, depending on what we buy, we can put up to 500 petaflops in that space, believe it or not. It's a power restriction problem. And so if you go to particular GPUs that produce a lot of power per watt, we always talk about bang for buck. Well, we got another one now. We also got to talk about bang per watt. And that's going to determine how much we can put in that second half of that data center. But on the best bang for what machines, which are a GPU machine, we can put up to about 500 petaflops into that. And that's proper petaflops, not cheating petaflops like Viridian. That's proper petaflops. So we can put about 500 petaflops into that machine room. What was the other part of the question, Ross?

That was both. One was on space and one was on density. That's great.

Yes. Thank you.

Okay. Thank you, everybody. I think we've run out of questions. Matt, thank you. It's very late in Houston at the moment for dialing in and Ajesh from Perth. This session will be recorded and posted on the website. Thank you very much, everybody, for your attendance.

See you later, everybody. Bye.

Thank you.