DUG Technology Ltd (DUG) Earnings Call Transcript

Okay. I think we might have everybody joined. Apologies for the late start. We're just having some zoom issues with Matt who is overseas. So today, welcome to the Doug Technology FY '25 results update and Webinar. We're lucky enough to have Matt Lamont online as well as Daniel to take us through an investor presentation where we will go through the results. Matt, if I could ask you to raise your hand, and I can give you speaking right so that we can get you connected and talking I can see -- I can't see...

I'm not sure if Matt had actually added on yet. So maybe, Steve, we can just get underway and then and then Matt's working with IT at the moment to get on.

Okay. So Dan, perhaps you can kick things off and run through the presentation. Would you like to share and drive the presentation?

Yes. Absolutely. Just let me get it. I'm just saying your request. That's showing for you now, Steve?

It is. Thank you. All good.

Great. Well, thanks everyone, and apologies for the delayed start. Matt is working his way onto the call, but hasn't quite met it yet. So he's over in Houston at the moment. We've got our big Image Conference, our biggest conference for the year next week. So I'm going to get started while we wait for Matt to join. So we really wanted to open up, I guess, with the key themes -- key themes that we've seen this year and through the result. And we think one of the really important pieces and certainly, we've obviously been talking to now since results last year is elastic multiparameter 4-way form inversion imaging. MP-FWI so what we've seen, and this will be something that Matt can give a lot more color to when he joins later. But what we're seeing is just outstanding results, and we're seeing that now convert really strongly into wind, and we've got a really a big backlog now with the pilot projects that we completed of really amazing results and examples from basins all around the world with our really key core customers. And so that technology now, every time we release a new algorithm, a new piece of technology. We've tested it to the end degree. But when you get it into a production environment, you really learn, you find the corner cases, you find the areas where maybe it doesn't apply quite as well. You hope it did. I just realized I've been looking down the whole time, but I've got my camera up here. And so the pilot projects and the production projects now that we're running with MP-FWI have been super positive because the results we're getting a blind every run away. And we're now starting to see that come through awards. And moving on to the second point here, we think the real standout result. The standout point in this result is the growing order book. And so we closed with an order book, here's Matt.

Well, I'm here.

Yes, you're all good. We're just on to the second point, but you probably can start over now that you're here?

Yes. Sorry, everybody. technical problems. But anyway, I've made it. Look, we're really excited in DUG. The next year looks fantastic. But in presenting today, we've got key -- three key themes for FY '25. The first is that we brought our elastic multiparameter FWI, which was a massive step forward because it is part of that whole replace your traditional workflow except that you went further, now you replace not just processing imaging, you also replaced your quantitative into Straight to Rock properties. That's been super well received. We've done a lot of proof of concepts now and they are converting into full-blown projects, which is very exciting. And that's really adding to our order book really nicely, the elastic. The order book of $52 million is amazing. And especially when you think that $46 million or nearly $46 million of it was added in just the last 6 months. So you can just see the huge momentum building up as we're going forward. And a lot of that is coming from our new offices, right? Abu Dhabi is winning projects, really nice projects. And Brazil, we won a really nice 4D project, a big project. And just to remind viewers 4D is the ultimate work that you want to get as a service company because it's repeat work, you repeat it, often you repeat it every 12 months because you -- as they deplete a reservoir, they want to redo the work to see how -- where they're draining the reservoir and where they're not draining it. So they do that drilling, infill drilling based on what you do in 4D. So it's a beautiful work because you get to repeat it every year. So it's fantastic that project came through from Brazil. Can someone change the slide for me, please. So what we do, just a quick recap for anybody who hasn't seen this before. Services is where we make most of our revenue and that's where our gives set and process data for a client. And that's where the Elastic & Acoustic MP-FWI makes most of its money at the moment, but we also do conventional seismic processing imaging, and we do multi-client as well, which we put under Geo services. Software is something we're super excited about. And we had some issues with where we registered wins, which took off a bit of shine when we dropped to 13% because of just an accounting at effect. But we're really excited, and software is really building up. And I think we're going to -- we're building momentum there, and I'm pretty confident that we're going to -- although we've grown at 13% this year, I believe it's going to grow faster than that in years to come, we will accelerate that. High-performance computing is very important for supporting our software business. It has been a bit disappointing in the last 12 months. But again, it's just a very small part of our business. And the emerging businesses, DUG Nomad and DUG Cool, we really look forward to talking about. Over the past 12 months, as we signed the DUG Cool deal and we -- I've been talking about it, we said we expected to make a -- we expected BAC to make a sale in financial year '25, but we didn't expect it to be material. And indeed, they've made a sale, and indeed, it's not material. So what we were expecting turned out. But Dan's going to address this later and they've put a lot of investment into this, and they're starting to gain momentum as well. And DUG Nomad, we've made our first Nomad sale. We're excited by that. The pipeline is building there. And again, Dan will talk about that shortly. Can someone change -- thank you. So this is the well according to DUG, whereas Orange is where we've completed projects in the past. But the really nice thing about this map is we've now got Abu Dhabi on the map. And that office is flying. The geo visits that we've got there are very good. Our fit-out of our new office is complete. We're just waiting for a little sign off from the authorities before we move in. So I think it's we're expected to move the week after next. So very exciting times Abu Dhabi. Abu Dhabi is a full blind office. I think we'll -- we've got seats in the first instance for 45 geos there. Rio de Janeiro is a different one. It's just more of a front-end sales office. We'll have a handful of geos there in order to satisfy local input local requirements, but it's more of a front-end office. So yes, it's really pleasing to see those offices on the map, and it's especially pleasing to see them kicking into our order book as they have. So financial snapshot. Amazing order book, $52 million and building, right? I expect that to continue to build. As we are saying, most of that has been added in the last 6 months. So it's 58% up on December 31, '24 and 42% up on 30th of June '24. So it's a significant increase in order book from last year. Revenue is disappointing. We talked about that down 4%, but it's about -- and I had hoped that we'd actually end up flat, but we ended up slightly down. So I was a little disappointed in that. EBITDA is down a bit, but we've been investing a lot of money into growing into people, into offices and all sorts of stuff. So yes, that is what it is. And as we said, the actual last 6 months has had an EBITDA of 30%. Services revenue down a bit as expected, software revenue up 13%. Look, software -- I keep talking about software when I'm talking to you folks, and I'm really excited by it. And we did increase our growth, but it was edge taken off it because of the way a few accounting things. But I expect that watch that number. That's going to -- we're going to continue to grow that growth number, I believe, strongly. And you see revenue, as I said, is disappointing, but it's a very small part of our business, but it's very important for software.

Great. Thanks, Matt. I'll take over for the financial slides. So I think, as Matt touched on, we really finished strong in the second half. I think that's really the key message here. In the first half, we were disappointed in, but the second half, we felt it was a really strong result. And so that was headlined by the $33.8 million revenue in second half and $10.2 million of EBITDA at our 30% EBITDA margin. the results, we think, are largely in line with consensus, and we think shouldn't really be a positive or a negative surprise for FY '25, but we think the headliner really is the order book number and the strong awards that have come through the second half. So to touch on that, we see services revenue down 5%, but we have $45.7 million of awards, which is what boosted that order book number. $45.7 million awards in the second half. And so that second half SKU meant that revenue is picking up. Momentum is picking up. But as the projects came in over that life, we expect to see kind of revenue growing as we move forward. It's important to note as well that of that $45.7 million, more than half of it is for MP-FWI work. And so it's been a really positive half in terms of solidifying the work that all the teams, the R&D team and all the teams have done in the MP-FWI and the direction that we're heading in the direction that we see the industry heading with MP-FWI. So to expand in on software and what Matt touched on, we had a few timing a few timing pieces where revenue that should have been in July last year was in June or revenue that was in June this year should have been in -- sorry, revenue that was in June last year than is in July. And that's what causes that kind of edge to come off. So just to bring it back, I guess, in first half, we did 22% software growth. In the third quarter, we did 23%. And then unfortunately, in the fourth quarter, we had some of these timing issues, which weren't lost sales. They're just timing issues with when the recognition occurs, that were picked up this year. And as a result, we ended up with the 13% for the full year. Now as Matt touched on as well, $15.4 million of EBITDA, that's 25% margin. We have 30% margin in second half and the significant investments, which we touched on at the half year as well. We had the release of E&P and the pilot projects that went along with that. And there was a lot of resources invested into those projects. And if you venture deep enough into our annual report and you see the segment reporting in there as well. It's worth highlighting that when services -- when our services business uses the computers, they pay a unit rate, they pay per server hour. And so when we're making these big investments, especially with Elastic, it's really pronounced when we're making these big investments and doing these pilot projects with even know or low revenue. That's where that swing comes in. And so services is making that investment proving out these pilot projects generating results and generating the social proof, so to speak, that we required to then go and sell and get the outstanding awards number that we got in the second half. But when we look at it on a segment level, we get services being unprofitable for the year, and we get HB saving, a step-up in their profitability. And it's not a comment on the profitability of our services projects or any issues around that and pricing. It's just a reflection of the investment and the significant investment that was made in getting those pilot projects closed out. So then closing it off, we had employee benefits up 6%, and this is largely driven, as Matt touched on, we've got the Abu Dhabi office now up and running. We've got a bunch of staff there, and a lot of those staff were hired early in the year. And so we have their costs coming through. And we have some cost savings, albeit we had some redundancy payments in first half. We have the cost savings coming through second half from the restructuring that we did in first half. And in D&A, we're seeing that step-up in depreciation, but that's just us seeing a full year now of the new compute assets and purchased in FY '24 half 2 and in FY '25 Q1. And so what we're seeing both through the finance expense and through the depreciation line is just a full 12 months of those assets being paid for. It's worth noting that the first batch of assets will be off the books, so to speak, in the end of January as well. So we're starting to get some of those some of those facilities coming to an end. And you'll see in the annual report, a more detailed commentary about this in the operational and financial review, where we talk through the way that we acquire these assets is using asset financing. And when we get technical with the accounting, they come through as a right-of-use asset and are treated like a lease. And what that means is that the tenure of the facility being 2 or 3 years is how we have to recognize it through the book. So we expect, we know that these assets will last 5-year plus. We have assets running that. We have computers running that at 8 or 9 years old, but we have to recognize them in an accelerated manner just due to the due to the accounting rules. And so that's what puts a bit of pressure on that DNA and on that finance expense line. So I think that's enough for that one. I'll move on balance sheet. So net debt of $3.9 million with total asset financing of $20.4 million. So that $20.4 million now is that number is coming down as we make repayments on the facilities that we have and the net debt of $3.9 million is down significantly from earlier in this year and last year as well. In October 2024, as many of you'll be aware, we completed a capital raising. So we raised AUD 31.4 million before costs. And just to remind you, even though I said we're reporting in all these numbers are in U.S. dollars. And so just to touch on use of proceeds for the data center infrastructure upgrades. The design is complete. We're now kind of in the planning and tendering phase. We've ordered a lot of the long lead time items. And so some things like electrical transformers and different items with the boom in the data center industry, we're looking at 40- to 60-week lead times. So we've had all those items ordered now so that we're able to move quickly and let this project kind of until as we move forward in the year. Worth noting still, as we touched on at half year that we have floor space available in that Houston facility. That's about 20%. But when we consider the density of modern computing, we've got 20% upside in our real estate. But when we look at kind of computer throughput, it looks like we could almost double the compute that we have today, if not more, with that 20%, and that's just a result of the greater density we can now achieve. Matt's touched on Middle East, so I won't touch on that one. And then DUG and Nomad first sale has been completed, and we'll talk about that a bit further on in the deck. Increase in other current assets. The only thing to call out there is when we do our asset financing, we pay a security deposit on the facility. And so that security deposit is equal to the last full payments on that facility. So we have some of our facilities, which expire at the end of January. And so our final payment for those will be in September with a small bullet payment at the end. And so that's what we see then coming through in the other line. Cash flow then, so cashless for suppliers was a result, same in the first half, actually, but it's a result of the third-party compute, that was utilized in the second half of FY '24 being paid off in FY '25 half 1. So that's where we get a bit of a step-up in payments to suppliers. But it's as a result of that third-party compute utilized in FY '24. Payments to employees fell, which is a result of the restructuring initiative and then the acquisition of -- that was -- the main part of that is the final batch of AMD Epic Genoa that we received in July 2024. And then we also had some initial waves of infrastructure capacity upgrades in Houston and the Abu Dhabi offers fit out as well. I've touched on the capacity, so I want to touch on that again. And then we have the USD 19.7 million, which is the USD figure from the October capital raise net of fees. So I'll hand it back over to Matt now as we kind of get back into the growth drivers and the business.

So oftentimes when folks look at DUG, I think that we're doing a number of sort of things that are sort of quite separate. But actually, it's -- the DUG ecosystem is complete. It's all interlocked with one another. Every piece is necessary for our business. The processing and imaging people we have, the teams we have are a major asset of our organization and split across all of the offices and are amazing. We've got a multiple library, which is in its infancy, but it's growing and will be a significant part of our business. The -- the software is fantastic software. It's for both imaging processing imaging and interpretation, including all of these. And we've got a big plan out for in how to grow that and how to move that forward technically and leading to growth. And I'm sitting in Houston at the minute. And this week, we've had so far, 6 visits where we put our plan forward plans to clients and to judge their reaction and see if they like it and they're genuinely excited from where we're going. I'll talk a bit more about software later. We have a lot of great different data analytics, different workflows, different AI stuff. There's a lot of stuff going on. And integrated with everything is that HPC backbone, which is all proprietary running our immersion called data centers, saving 51% of the power bill really important for the environmental impact. And often -- and I'll run over one of the questions people get put to us is why don't we use the cloud. So firstly, the thing to remember is that the cloud is not just there for us to grab. So we -- for us to get what we need, we would have to sign up for a number of years. So you don't get flexibility in how much you use you get flexibility, if you go into the spot market, but then we can't run our business to the size we are in the spot market. So that's the first thing. You don't get that flexibility that people often think comes with the cloud. And the next thing is, and Dan has already talked about it, is that even when we're making repayments on the hardware, we're cheaper than what the cloud would be, but once we've paid off, like we've just paid off the first batch of new compute, once the compute is paid off, then we get a free run of it, right? And as Dan said, we've got machines ringing that are 9 years old. So if we pay them off after 3 years or 2 years or whatever, we've then got 6 years of essentially free compute, just paying for the consumables. So that HBC and the way we manage it and the savings we make is super important for our business. And it's a major advantage over any competitors that use the cloud, and I seriously don't know how they come to the decision to do that. And then the emerging businesses on Nomad and DUG Cool, which Daniel will talk more about in a minute, are really important for where we're going, and we're excited by them. And it's just been really, really great seeing BAC that after that business. and Dan will talk you through what they've done, but they've made some serious investments into the future of that business. We keep talking about this, and it's exceedingly important. We are leading the way with seismic imaging in the world. Our Elastic MP-FWI imaging is the talk of the town. Our big -- we've got 2 big global conferences each year. I'm over in Houston because is the Americas one is next week. But earlier, in early June, we had the big European one, which was in Talus. And all the majors, the super majors, so your shelves, your Exxons, your total even, at BP and so forth or gave talks and their talks were on MP-FWI and how that is the future. And so all the talk now is about MP-FWI,and they've set the edge. They -- although we came out 2.5 years ago, it's taken a long time to get to this point. They are now saying this is the future. And one of them actually even met and said, DUG is on the right track, but nobody else is. So that was a great rate endorsement for us and really helped us with clients. So now we're coming on board with oil companies that have never worked with us much in the past that they were 100% Viridian shops, and we just won another 4D project today from a company that's never worked for us before because back of that MP-FWI technology. So momentum is built and is exciting. Going Straight to Rock properties is extremely exciting. That's a massive step forward for the industry as a whole. And in the multi-client, multiline business is a really good business. It's high margin. It's basically where we own the seismic assets and then we license them to oil companies. And so rather than processing data and being paid once for that, you actually on the seismic data and you license it out. It's a really good business. It's in its infancy in DUG, but it is growing and watch this space. Our software is the leading software in the industry, I believe. It's certainly the most modern and leading software amongst our key services competitors. It's -- the software business itself has got accelerating and it's going to continue to accelerate that growth. It is recurring revenue. We we get money up front every 12 months. Except for some of the software that runs on the cluster, if they run on the cluster in -- there's -- they pay for the hardware, but they also pay a little fee for the software that they're running on the cluster. So there's consumption-based billing for the cluster side of it, but there's -- most of the money is an upfront license fee. He's a one-stop toolkit, which is what everybody wants. I think it's the only one-stop toolkit out there that goes from processing and imaging right through to multi-parameter FWI, Rock properties, Rock Physics, inversions or within the one package. We've been selling the process imaging software, as you would be aware for the last 4 years, and it is really growing. It's a great area. Although that you might think that it would cannibalize our services business. It's probably number one, we love -- we really love the software but for obvious reasons. But number two, a lot of the small service companies that are picking up the processing and imaging software actually work on projects that we wouldn't really like to work on in services. So we have a great client in Pakistan, who works on a lot of projects in the stands and in Pakistan, and they tend to be small projects and cheaper projects, not projects that we'd like as a service business, but it's great to get a revenue stream from those projects nonetheless. And then another example I've used in the past is, again, is a company called Rock Wave in London, and they do a lot of wind farm seismic work. Again, the projects tend to be smaller and not really what we would like in the service business, but it's great to get a revenue from them. We have a big plan for the software, as I've alluded to, we're presenting that plan this week and next year to clients here in Houston, and it's been really well received. And we really -- the whole plan is -- at the minute, the software is used by our competitors for their multi-client businesses, so like Viridian and TGS. It's used by small oil companies. It's used by a lot of consultants, and it's used by some medium-sized oil companies, but not generally throughout their business, just in some groups and teams. Our aim is, therefore, to grow that share of those medium oil companies and to grow into the larger oil companies and even the super large oil companies. So that's what the plan is about. That's what we're presenting and getting feedback on and the feedback so far has been great. But that's how we're going to grow this business.

I'll pick it up. So DUG Nomad and DUG Cool, our emerging businesses. So for those who are not aware as well, DUG Cool, it was an internal invention that we had patented in 2016. Ultimately, we ended up building out our Houston data center, and we won Data Center of the Year, an award from a publication called DCD. And that just got us some notoriety. And it started meaning that we were getting inbound. And so as we were getting inbound from companies who are interested in the immersion cooling, we started to work to pursue it, but realize that we weren't really set up to do it, and it wasn't our core competency. And so what we did in August last year, was we signed an agreement with Baltimore Air oil company, and they've really since then taken that and started running. So BAC has rebranded what we would call DUG Cool. They've rebranded at BAC COBALT. And they've just finished or they're just in the process of finishing their new research and research and development hub for immersion cooling in their global headquarters in Baltimore, which is a really exciting development. They've got a full team working on it now. They've got teams focused on building out their partnership in this kind of computing ecosystem, whether that be the fluid or the servers or all these different components that go into a tank. And they've also got a team that's been on the ground selling, getting feedback. And they're also doing R&D. So they've got their team there doing R&D. They've been doing it now for quite a while, and they've made little tweaks and they're getting improvements in their understanding, and it's been really exciting, how they've been going. So as Matt touched on, we had our first sale not material. But just in case anyone's wondering, I guess, where that's going to come through the P&L for the time being, we'll be putting Nomad and DUG Cool revenue through the HPC line we'll be sure to give detail and talk about that in detail, but we're not going to be looking to add lines and lines and lines to the P&L at the moment, and we'll be putting this through the HPC line. So DUG Cool is going really well. DUG Nomad also going really well. So in June, we announced our first sale, that was to Perstorp, which is actually a moment to BRB and Perstorp, they all subsidiaries of Petronas. So Petronas is the end client, the units being delivered into Malaysia, and it's going to be delivered during this first half of FY '26. So we've not actually got any revenue through on that, albeit we've received the cash deposit, but it won't be recognized as revenue until later in the year when it's deployed. The pipeline there is just continuing to build. We've got some key conferences coming up in September and November, where like this image, you can see in the top right where we'll have the DUG Nomad at on the berth. And those are really key touch points and really key BD times for us. And we're continuing to refine how we sell message and continuing to learn about that business and make sure we're getting about it the right way. So it was really great to get the first sale of good in June and now the really healthy pipeline there that we're working to get over the line and we're really excited about. The difference really with DUG Cool to touch and this applies both across the BAC COBALT system. And DUG Nomad is it just simplifies the system. Our patent simplifies the system by putting the head engine into the tank. And that allows you to do a few things. One is it allows you to be super debt. So in that 10-foot container, which you can see on the top right, the section you can see there, we've got a tank. It's one of our normal tanks. It's got 2 heat exchanges and it can do up to 80 kilowatts of IT heat rejection. We think we can get that number seven, but that's what it does for the moment. And behind that, what you see is a chiller. And so the cost the idea here that it's been really well received is it's a 10-foot super mobile data center. So you can put 80 kilowatts of IT, you can put it wherever you want. You don't need external cooling or your net is to add the power that's been really well received. But -- and we've got this big pipeline we now just need to get it converting and make sure that we're continuing to learn about the messaging and the way to get those deals closed.

And so you're aware that we've been globally expanding. I've touched on how well Abu Dhabi is now going and how well Brazil is going. So we've really now got our offices in the key locations around the world. The only place we might put a front-end office like we've done in Brazil, might be India and maybe Vietnam. But I think we are really -- we are -- and we're winning a lot of Vietnam's work as it is. So I think we are certainly major offices. We've got all the major offices we now need to cover the world geographically. And it's just fantastic to see Abu Dhabi in Brazil kicking into the revenue and to the order book, the way they are. The highest ever order book, and I think you -- that's just going to continue to grow from what we're seeing. We've talked about a really large, healthy pipeline of biggest we've seen for the last 6, 12 months, and now we're starting to see it really convert over the last 6 months. But that pipeline is continuing to grow. It's not dwindling. It's continuing to grow significantly. And you can see there, half of that order book came in the last 6 months, not a half -- sorry, $45 million out of $52 million came in the last 6 months. So momentum is seriously growing. Yes. And we're just -- we're continuing to say scaling and the data center, the next to last next to last fit out of the data center is complete. And so we've got a lot of space capacity there now ready to go for more compute, should we need it. and the long lead time items have been ordered for the remaining piece of the data center that's not complete. And that's 12 months out now. To get those long lead time items, they're very slow at the minute because of all the data standard bills around the place. And so the rest of the equipment to complete that will be ordered in order to in order to complete at the same time as those long lead time items come in. But we've talked about services. Can you go back, please? We've talked about services benefiting in Abu Dhabi and Brazil. What we haven't talked about is that we are looking for and we'll be adding salespeople for software into those offices. They're not there yet. And we'll be adding Nomad front ends and so forth. So everything else is going to follow on now from services in those new offices. So finish off. Why invest in DUG? So we're a founder-led innovative culture. But within the realm of what we're doing, we're not out looking for new businesses in anything, but we have the absolute industry-leading Elastic MP-FWI, that's been totally acknowledged now in the industry. We've completed 75 projects now, which blows some people's minds when they hear that. We have what I think is the best software in the industry, and it's certainly the broadest going covering from processing imaging all the way through to Interp and Rock physics and so forth. We've got the painted immersion cooling, which is now being licensed by BAC, and that's just going to ramp up and get bigger and bigger. And DUG, no, Matt, it's a slow start to no matter, but there's a really good pipeline and we are learning about how to get after that business. We got a great global footprint now. We've got the global footprint we net we don't need to open up any more full-blown offices. Middle East and Brazil are really kicking into the coffers and the order book now, which is fantastic. The Middle East didn't kick in any money last financial year FY '25. The first invoices went out in July, which was fantastic to see in a real milestone. And no invoices have gone out from Brazil yet. We've won the big project, and I think the data has arrived and so we'll start loading data, which means we'll start to work on the data and be able to invoice shortly. So really, really great to see those offices really hit their stretch. Brazil hit its straps very quickly. And Middle East has been a lot of work done on the ground there, and it looks fantastic. But the pipeline there is quite staggering, to be honest. We've been at it for quite a long time, now 21 years. And over that 21 years starting in a grain Perth, we're now globally recognized -- globally recognized as a technology leader -- we've got a great footprint in the industry, and they're just really well known and respected. We work for everybody that you could think of. It's a case now of working our way up and doing the most more 4Ds and more technical stuff for the big guys, but we work for everybody and significantly. And in terms of momentum, [indiscernible], we've got a massive order book. We've got the biggest pipeline I've ever seen by miles and it's continuing to grow. And momentum is building and the software business, I believe we can accelerate that growth even further from year to year for years to come. And if that's not all enough, you've still got these emerging businesses of DUG Nomad and DUG Cool, which are just starting to find their feet. Thanks. Steve?

Okay. Thank you, Matt. We might move to Q&A. [Operator Instructions] If we can go to Jack Daley from Shaw, please, to ask the first question.

You can hear me, right?

Yes.

Congrats on a great result and great to see order book up $52 million I guess just in terms of the kind of quarterly contract wins, so it was $23 million this fourth quarter, about 22.5% last quarter. I guess we're 2 months into this current quarter. Is that and you're looking at the demand that we're seeing, is that kind of the run rate that we should be thinking about that step up to like the mid-20s -- low 20s contract at once third quarter?

It will be -- it will still be lumpy, Jack. I would -- some quarters are always stronger than other quarters, and it changes a little bit, although when we look at wins and proposals written and so forth, we always compare them with the same month from previous years and life for writing proposals. And so you get that and then you also get some -- if you write $20 million worth of proposals 1 month, and you expect to get a lot of wins in 2 months' time. So I would rather than looking at it on a quarterly basis and say that's what we would expect on a quarterly basis, I would probably look at it on an annual basis and say, this is what we expect on an annual basis but growing from that.

Okay. And I guess it's kind of the second half is probably indicative of what you'd be thinking about from an annual basis?

Certainly and growing. I expect it to grow, yes.

Okay. And then I guess just on the 4D work that you spoke about seems really exciting and especially the repeatable nature of it. Are you -- can you give a sense at all on like the quantum of that? And is that something that you're going to be selling more moving forward?

Yes. We try to sell it. It's the pinnacle of seismic processing and imaging, so it's the hardest work to get -- but yes, we've got -- we won another one, as I was mentioning today. And -- and the quantum of that project from Brazil is very large. It's in the millions and millions of dollars -- multiple millions of dollars. So it's -- and as I said, it's 4D, so you would expect it to be done regularly. Sometimes, they're done every 12 months, sometimes they're done every 18 months, so forth. So you expect it to. So that's absolute. We have -- we've been doing a great 4D project. I think we've repeated it 3 times now for Chevron in Western Australia. And yes, so we're certainly chasing that work.

And I guess maybe just last one. I think maybe when we think about kind of the size of the opportunity in Abu Dhabi and Brazil, like you did for the year, $10 million in Malaysia, $28 million in U.S.A. and $20 million for the U.K. What kind of profile should we be thinking about these geographies to cover in the short and medium term?

So I think Brazil itself will be -- is part of the Houston business unit at the moment. So it will add to Brisbane. It led to Houston's revenue. But the quantum out of Brazil that we're seeing Well, we'll see. It will come in somewhere between -- I think it will come in somewhere between APAC and London, somewhere in that $15 million worth is probably what I would expect it to do this coming year, something around there. The Middle East, honestly, could be anything. Some of the projects we're looking at there are just so large. You win one of those. And well, you're away to the races. So it's very difficult to predict in the same way just because it's -- some of the projects are so large. But Jay has done a tremendous job there. And we're known. People who know us, people know Jay. We're really getting involved in a lot of tenders and opportunities and stuff. It's amazingly slow to get things done that things just take a long time. So but we're getting used to it. But once it's rolling, of course, you're up in the way.

Allan Franklin at Canaccord Genuity, if you could ask your question.

Have you got me now?

Yes, we got you.

Well done on the progress. Just wanted to clarify 1 or 2 things, Dan, just on the finance cost of things around the compute, just to sort of clarify if there's any nuance on the cash flow side of things versus P&L Because you are sort of thinking some of the cash cost of that compute starts to roll off in the current half?

Yes. I guess when you're looking between the P&L and cash flow, it's probably worth remembering that those right-of-use assets, they come through the depreciation. So our repayments come through depreciation and finance which is not always traditionally, if you're talking about CapEx or depreciation, it's not always necessarily a cash cost, but in this instance, with those facilities as it is. and so that can cause -- it's worth keeping in mind, I guess, when you're in between the P&L and the cash flow and trying to tie the 2 together. I'm not sure if that's what you are cutting on.

Yes. Look, correct, yes, maybe some of the compute was obviously received later in the period as well, but just noting, I think F '25 was probably the peak drag on cash costs from the compute. Just in terms of the customer concentration. I know you haven't sort of drawn it out in any detail, but that looks to have changed, certainly so for your larger customer being less of a concentration during F '25. If there's any color you can provide on customer concentration, please?

Yes. I think it's definitely come off. so that we have some really big key customers do, but we don't have one customer that's 20% of revenue or to that effect where in years gone by when we used to do that graph in the deck. You would see that kind of one kind of dominant customer. We don't really have that anymore. And so we've got a good swath of really important significant customers, but the concentration in one particular one has come off a bit. And that's just due to the cycle of their projects. It's not -- we're still working for them. We're still doing all their work. But they just have a bit less work at this moment as they work -- as they get through the work that we've done for them over previous years.

Yes. And I probably don't necessarily order at a group level, but just intrigued by some of the sort of divisional cost allocation changes because it does look like that's shifted around a fair bit between divisions during the period?

Yes. But the method is actually largely the same. I think what we're seeing in FY '25, the biggest shift is we when we're allocating a lot of the costs, not all of the costs, but we're allocating a lot of the costs, especially corporate things that run across the business. We'll do it on a revenue basis. So if take HPC, for example, you might see a step-up in revenue based on what we touched on earlier that we've got the new compute online, and there was a lot of resources invested into the pilot projects. And so then HPC is generating more revenue as a segment. And so then because it's generating more revenue as a segment for the costs that are split based on revenue, it's then getting a higher allocation from that pool. And so it's not that the base cost of operating our infrastructure has stepped up. We have some higher power costs because we've got more equipment running. But it's not like there's a fundamental step change in the cost of running those facilities. It's just because the unit is making more revenue. They pick up a higher proportion of the corporate kind of corporate costs.