Home / Transcripts / Gorilla Technology Group Inc. (GRRR) · August 24, 2026

Gorilla Technology Group Inc. (GRRR) Earnings Call Transcript

August 24, 2026

NASDAQ US Information Technology Software earnings 45 min

Earnings Call Speaker Segments

Jayesh Chandan executive
#1

Thanks a lot. Good evening, everybody, and I appreciate you being here. For me, the opening six months of 2026 represent a really pivotal leap ahead for Gorilla. Sales climbed 99%, close to 100% versus the same period a year earlier, rising by $78.4 million — in effect doubling across that initial 12-month stretch. Even more significant, though, is that momentum kept building through the half. Q2 revenues came in well above $50.1 million, a net gain of about 78% over Q1 and 138% compared with Q2 of last year. Initially, as we'd pledged to the market, we were looking at roughly 33 million, and we raised that guidance to $44 million. We beat it by an additional $6.1 million, or almost 14%, which mainly reflects that every deliverable and certain milestones got finished ahead of schedule. To me, that is exactly what execution looks like. The other challenge we faced is that operating performance and progress were, at the same time, just as meaningful. Our reported operating loss shrank from $41.1 million in Q1 down to about $2.2 million in Q2 — a 95% decrease. A large portion of the first-quarter figure came from share-based compensation that had already been recognized, and over 80% of the H1 share-based compensation was taken up in Q1, while the quarterly charge dropped by roughly 78% in Q2; in < >, Q1 bore the vast bulk of the weight, and Q2 presented a far clearer view of the operating momentum we required; our cash efficiency likewise got much better. Even as revenue rose by about 100%, operating cash consumption dropped by roughly 65%, from $12.5 million in H1 of 2025 to $4.3 million in H1 of 2026. Operating cash usage also went from 31.8% of revenue down to only 5.5%; these are hugely material numbers that we must factor in; the company additionally posted a total cash increase of about $179.8 million across the first half, largely reflecting financing support and the expansion program, along with customer collections. So we closed June with around $179.4 million in cash, roughly 82% higher than our closing; that capital isn't simply parked in a bank earning some interest rate. We put it to work. What we're doing is buying infrastructure, locking in capacity, getting sites ready, building teams, and funding the deposits plus working capital needed to deliver projects of a scale never previously undertaken. And just FYI, right now we're preparing about 5 separate sites simultaneously, and that demands human effort. At the same time, we recognize that cash balances will shift between reporting periods. Investors ought to tell apart cash being burned by an underperforming operation versus capital being intentionally deployed into contracted projects and revenue-generating infrastructure. Those two things are not remotely alike; however convenient it might be for some to suggest otherwise. This investment stage, though, is also the reason behind the gross margin picture we're seeing right now. Our gross margin reflected a revenue mix that leaned heavily toward hardware, along with initial deployment and project mobilization. What I need to be sure of is that Gorilla has also put more than $14.1 million into property and equipment. Right now, that figure stands at $29.4 million. We're also building out the installed base as infrastructure gets commissioned, customer workloads migrate, utilization rises, and we anticipate the revenue mix will broaden toward compute, monitoring, managed services, and every other associated service. Hardware, guys, honestly doesn't start generating its complete financial results the moment it leaves the factory. It has to be delivered. It must be installed, it must be powered up, it must go through testing. It has to win customer acceptance, and even more importantly, that's when utilization kicks in. Even more importantly, we want to be certain we're moving extremely, extremely fast. Now, on updates. I believe the market has been pressing me for updates for quite a while; take Yotta Phase I, for instance — testing there is finished. Equipment deliveries are in progress and deployments have started. Testing will begin by the close of this week or early next week. For Yotta Phase II, the equipment is being manufactured right now, with completion anticipated within the coming 25 to 30 days. In Indonesia and Batam, we're collaborating very closely with our OEM and infrastructure partners; as everyone is aware, we have secured substantial data center space, and we're aiming for roughly another 200 megawatts of capacity, with an initial ready for service in the middle of 2027 and the broader deployment anticipated in the second half of 2027. At < > in Thailand, the land has been cleared, and we're progressing with the financing, the GPU procurement, and the infrastructure requirements while engaging prospective offtakers, with the goal of shifting into discussions on firm customer contracts. Now, to be absolutely clear, megawatts and adjusted capacity are not — they aren't simply capacity sitting there. Even more importantly, that capacity has to be commissioned, it has to be contracted, and it has to be utilized. Now, our accountants stay very stubbornly unwilling to accept electricity as a payment. So unfortunately, we have to ensure customers pay us, I think at a given point in time. Alongside these major infrastructure programs, our established security and network intelligence operations remain a vital part of Gorilla. They deliver not only the customer relationships. They also deliver a better delivery experience and cash collections that underpin our broader expansion. Now, for those who think we're pivoting — we aren't. We're not abandoning the business that got us here. We're using it as a platform to build something substantially larger. Now, for Q3, we're planning revenues between $48 million to $50 million, compared to the previous plan of $36 million to $40 million; that represents a significant jump, roughly 20% to 39% above the earlier planning range. Regarding Q4, the plan we're operating under calls for sales coming in above a range that comfortably clears $60 million to $70 million. When you add in the $78.4 million we brought in during H1, the guidance we're now giving for 2026 sits at no less than $200 million, an increase over the $137 million to $200 million band we put out at the start of this year. Hitting the top of that band calls for more work on our side, such as additional deliveries, plus customer activation efforts. We fully plan to go out and earn that top end. I'd like to ensure we aren't just putting out an announcement — rather, we want this to grow increasingly valuable quarter by quarter; moving on to look further out at 2027, the goal we've set is roughly $450 million to $500 million in revenue. That's a bold objective, equal to quarterly sales of approximately $112 million to $125 million — sorry, roughly 120 — $112 million to $125 million. We don't count on the calendar to hand us that growth. Reaching the goal hinges on all the capacity currently being brought online, the commissioning of further projects, the transformation of pipeline demand into signed contracts, and shifts in how much of customer workloads get utilized. Plenty of critics have been sounding off from the sidelines, unfortunately. Sadly, that doesn't get GPUs delivered for us, nor does it commission data centers or bring in dollars from our customers. Execution is how we answer all of that. Gorillas came into the second half with revenue scale that was substantially larger. Quarterly performance that improved dramatically, stronger liquidity, and an expanding lineup of major international projects. There's still more work ahead; we're keeping absolute delivery discipline. We're handling capital very carefully while utilization improves, turning opportunity into recognized revenue; and make no mistake, the trajectory we're on now cannot be mistaken. So we've moved past explaining what we're going to become — we're starting to show it. Thank you very much. Bruce, over to you.

Bruce Bower executive
#2

Thanks, Jay. Jay covered most of the key points, though a few things deserve more detail or emphasis from me. Let's start with the top line for the first half: $78 million in revenue, up 99%. You can see from the guidance numbers that when it comes to revenue, Gorilla is in hyper-growth mode. These results make us pleased, and we anticipate this type of growth trajectory will carry on. Beyond that, as Jay noted, gross profit in the first half reflected both an investment into the business and the mix — and as that mix gets better, we anticipate gross margins expanding. There are a couple of ways the mix will improve. To start, when Yotta and the remaining GPU-as-a-service initiatives launch in the second half of the year, they'll deliver gross margins at or above 75%. Obviously, the company's total gross margin will widen because of that, and from that point it carries down to the operating line — significant operating leverage will emerge from it as well. So it will likewise carry into the underlying EBITDA figures; we've already noted some specific expense items. I'd like to flag that this was no ordinary first half. There was a notable impact from currency — sizable shifts in foreign exchange rates everywhere, which came about because of the run conflict, freedom day tariffs, and the tariff wars. This matters a lot for us since we hold a large amount of exposure to foreign currencies throughout the Middle East and Southeast Asia — the two regions hit by both geopolitics and tariffs. As a result, our underlying figures swung around considerably. Now that things appear to have settled, we think some of those below-the-line expense items we recorded in the first half will calm down too. On top of that, Jay mentioned a stock-based compensation item. We've already shared this in every filing we've made, so nobody should be caught off guard by it. I believe that's now in the past, so looking at it that way, the P&L for the second half should be relatively uncluttered. There are a handful of balance sheet items I'd like to point out. To begin, as we closed out the quarter — transitioning from the first half into the second — we had $175 million in cash; our conventional debt, meaning the kind from traditional bank loans, stood at $13 million, and on top of that came $107 million in long-term debt tied to a 5-year convertible we placed in June. We also finished a convertible in July. As a result, we have truly ample cash going forward; right now is the stretch when we're paying for equipment, for equipment deliveries, and for getting it set up. That will convert into sales during Q3, and then to a substantial degree in Q4. When we did that initial rollout, we paid for those units out of pocket—drawing on our own balance sheet and capital—and afterward via a small financing arrangement pegged directly at the project level. Now with Yotta 2, we've again put down considerable advance payments using our own balance sheet. As for what's still owed, our intention is to fund it via a larger facility with the project client, and I'll share additional details once I'm in a position to do so; however, my sense is that all of you on this call will be very happy with the terms we managed to get. Now, a handful of other items. On the investment side, as Jay pointed out, you're already seeing it show up in CapEx. The first half saw $14 million go toward that, and the second half is naturally going to be many, many times bigger. Going forward, that's where spending on the business will appear, and it fuels both revenue growth and a bit of margin expansion. Despite how heavy CapEx has been, I'd argue cash flow at the business is actually holding up well. For the first half, operating cash flow landed at minus 4 million, which is a narrowing compared with minus 15 million during the same period a year earlier. My expectation is that those operating cash flow numbers will continue improving. We're looking at a lot of money coming in from customers during September and October — over $20 million. With that, SG&A and the other operating expenses throughout the business won't be overwhelmed, so for the entire year we should end up at breakeven or with positive operating cash flow. One last thing I'd like to point out: you can find the guidance figures in the press release. For this year, our guidance is $200 million plus, and for next year, revenue guidance sits at $450 million to $500 million. Our approach to guidance is to use contracted revenue for which we have both an amount and a date. When a contract exists but the schedule or the dollar figures haven't been finalized, it stays out of our guidance. Our approach, again, is to set expectations low and then beat them in the marketplace; we prefer to base our commitments only on what we're confident won't disappoint. So when Jay walks through the delivery timelines for < > and the subsequent trip to the project in < > — that's what we call it — since those timelines are solid, they're reflected in guidance. For timelines that remain uncertain, guidance doesn't reflect them, but stay tuned, because as those delivery schedules solidify, our guidance could shift accordingly. At this point, we're not providing any gross margin projection for the coming year; honestly, it really depends on the timing of when those projects begin. Even so, I want to remind everybody that with Yotta 1 and Yotta 2, we're looking at an average gross margin of 75% on those. So if we're talking about revenue between $450 million and $500 million, the gross margin from that would be largely—most of it is going to come from GP as a service; therefore, my expectation is that you'd be seeing gross margins somewhere in the 40%-plus territory for next year. Once again, that isn't formal guidance, it's simply running the numbers. On the financing side, as I noted, there's already a financing offer sitting on the table for NutriDC. Beyond that, we've also applied the convertible's proceeds toward the first portion of that. When it comes to Yotta — and Yotta 2 as well — we've directed essentially the entire convertible proceeds to cover prepayments or deliveries, and the remaining deliveries are being funded through two separate project finance facilities, which we're either in the process of arranging or already have arranged. So once more, I believe that going forward, when you look at the company's growth and CapEx needs, you'll notice a greater emphasis on project finance or debt in general. We'll share additional specifics on how that operates in the next few days. With that said, I'll hand things back over to Jay.

Jayesh Chandan executive
#3

Thank you very much, Bruce; we're now open for questions.

Brian Kinstlinger analyst
#4

A great deal has gotten done, and there's a great deal worth talking about. Bruce, you raised guidance — at least for the upcoming year — and any year depends on what's already locked in, which means you know the sums coming in, the projects at play, and when they land. So turning to 2027, what's on the delivery timetable, which phases will be active, and which programs? Does it span all of Yoda's first two phases along with D.C.'s phase one? Or maybe take us through what's baked in versus what isn't?

Jayesh Chandan executive
#5

Correct. So on the first phase of the Yotta contract, we have a delivery schedule in hand and are taking in the equipment right now; so that's assumed, and it comes to roughly $100 million annualized in incremental revenue. Next is what we call the T2 contract — it splits into two separate batches. For batch one, a delivery schedule is in place, so that's assumed, and it totals $250 million in incremental revenue. Last is the < > project; in the initial press release we mentioned a deployment expected of 300 servers around the September timeframe, then another 700 servers by year-end. The delivery schedule for those 300 has firmed up, so it's included in guidance — that's approximately 30%, taking it to about 260, which equals an incremental $75 million to $80 million annually. The rest then comes from existing contracts, or ones we've already won and you've seen announced, and that's how you get to the $450 million to $500 million mark. So the 700 servers fall under TDC's first phase and aren't in guidance. TDC's second phase — 875 servers — also isn't in guidance, and neither, right now, is part two of Yotta's second phase. Of course, we'd want that delivery schedule nailed down. GPUs are like spice — they're an incredibly valuable commodity — so having them in hand beats talking over schedules, timing, and amounts in advance. That, again, is why we take the ultra-cautious stance of 4 million in guidance, with $260 million sitting outside it.

Brian Kinstlinger analyst
#6

Great. And then a follow-up < >.

Bruce Bower executive
#7

Sorry, Brian. So from our standpoint, that confidence comes from the capacity and the projects we're currently putting in place, like, as Bruce put it so well, right, we've got a delivery schedule for Yes. Actually, there are a whole bunch of them on the plane at this very moment. They're in progress. We're kicking off all our deployment, which basically means the power drops exist, the connectivity exists, and the Phase II equipment is finished as well and will be roughly delivered — deliveries are going to begin arriving. So we have a set time line. That means the first step has to be delivered by us before this month ends. The second set in the middle of next month; and then — so from now until the end of November, Yotta 1 and Yotta 2 Phase I are going to be complete. And then there's the big elephant, which is the < > that one specifically, actually, the initial 300 servers must be deployed before October ends. Testing is probably going to take 20 to 30 days. So we're looking at becoming operational at the end of November, the first week of January — December, sorry. Now beyond that, we were also working toward closing the rest of the 700 at the 875 service, which obviously — once that's finished, we will definitely revise the new targets for next year. But again, the difficulty we face today, just to tackle what the difficulties are, GPUs are scarce. CPUs are scarce, all memory and storage is scarce. Then there's the tailing and everything else. Beyond that, electricity appears to be somewhat of an issue. But what we're doing is ensuring that everything we've committed to is tied to the operational milestones we already have in place, whether that's equipment, capacity, customer contracts signed, workloads migrated, and all of the utilization being carried out.

Brian Kinstlinger analyst
#8

Great, that helps a lot. Yeah, one more question. As I recall, the NeutriDC facilities were already constructed, if I'm remembering right; but < >, is there any building work that needs doing?

Jayesh Chandan executive
#9

No, not at all. Yotta is an amazing builder since they construct everything on their own. So they've handled all of that themselves. And incidentally, every one of the Yotta data centers is completely prepared for service. Every floor is prepared for service. What we're doing at this point is just bringing in the power crops and having the GPUs undergo testing.

Bruce Bower executive
#10

Alright. So, regarding the project financing from the prior quarter — at that time, you said you had offers, and you put comparable offers forward, and your remarks now sound similar. Brian, what's been the main hurdle when it comes to getting that project financing locked in? Let me be very precise here. I never stated that we had offers. What I said was that in one instance, we have a range. And in the other instance, we are near the finish line. The hurdles have been, number one, our own push to obtain the most favorable terms we can. Some early offers weren't in shareholders' interests, so we responded, no, we will not take that. Then, number two, more lately, there's been a change affecting some of the delivery schedules. That change, in fact, involved them wanting us to pay — the vendors wanted faster payment from us — so we had to get people moving quickly to align with different schedules, and as a result, we do certain things. But overall, I'm quite pleased with where we stand. I'm quite pleased not only regarding the project finance situation, but also regarding how we are positioned in the debt markets as a whole. And then, as we mentioned before, we are in the process of pursuing a credit rating, so keep watching for news on that.

Michael Latimore analyst
#11

Jay, Bruce. Appreciate you bringing me in. It's encouraging to see other areas progressing nicely. I had a question regarding what the GPUs are worth at the end — do you plan on offloading them once 5 years pass? If that's the case, what kind of residual value are we talking about? And separately, could there be a chance to continue running those GPUs into a sixth year?

Jayesh Chandan executive
#12

Really good question. Great to hear from you, and hopeful too. So yeah, based on what you'd find across different sources right now, 300s and D300, at the close of year five they'd still retain value. You're roughly talking somewhere in the 20% to 25% bracket. That said, that's as of now, and nobody can predict what things look like 5 years down the road. Ultimately though, we intend to keep those GPUs in operation. We'll also — offers have already come in, whether from institutional investors or data center operators, basically saying, hey, I'd be glad to enter into an agreement with you once year five wraps up, we'll take it, and we'll have it appraised by a top-tier firm such as < > or PBC or KPMG; after that we'll assign a value to it; but the gorillas intent is to keep going. As the models develop, you're shifting from training workloads over to inferences, edge is going to become a significant sort of — we'll be making a significant move around the start of next year, we're looking at large-scale deployment across these regions. Another factor worth considering is sovereign AI — every one of these nations where we currently operate, be it the Middle East, portions of Europe, portions of Asia, they're aggressively seeking GPUs. They want to lock in their future, particularly governments and so forth, and we can deploy at that moment. So all things considered, either it's incremental revenue-based work extending through the following 3 or 4 years, or there's instant liquidity from selling to an existing data center provider.

Bruce Bower executive
#13

I'd like to jump in here as well, Alex, if that's alright. From an accounting standpoint, we're going to depreciate this equipment across a 5-year period. That's the lifespan we're assuming, mainly because the bulk of the contracts we're entering into run for 5 years. Still, I want to stress a couple of things: first, as Jay pointed out, there are multiple paths to monetizing these assets down the road. Beyond that, I'd note this topic has come up many times before — namely, how long can a GPU actually remain useful? Certain competitors out there have now had A100s operating for 7, 8, even 9 years, and those units continue bringing in revenue along with quite solid yield on cost. Naturally, A100s don't command the same rental rate they did 5 years back, yet the return remains quite healthy and the economics stay very attractive. So everything we're seeing suggests the actual service life could end up being considerably longer than 5 years.

Michael Latimore analyst
#14

Great. One more question regarding your CapEx projection for India and Indonesia — does that figure cover maintenance CapEx, or does that come on top? And if it does, what's the right way to think about the timing, or the cadence, of that maintenance CapEx?

Bruce Bower executive
#15

Yeah, I'll handle that one; so the CapEx guidance we provided is for the upfront CapEx, and upfront really means networking equipment, GPUs, servers, cabling setup — everything required to get things up and running. As for maintenance CapEx, we treat that as an expense, and there are a few reasons why. First, the majority of it is actually going to be labor. We run our RMA service, where we've got people who will service the equipment and also repair it. Second, a lot of that equipment is going to be under warranty from the vendor. So we'll spend a small amount to fix things or swap out spare parts, and so on, but the large CapEx that would be required if a machine breaks or a whole server breaks wouldn't fall on us. I'd say it would be with the vendor. So essentially, when you look at the forecast and at our gross margin, we include in gross margin all the spare parts cost and labor cost; it isn't broken out separately as maintenance CapEx.

Michael Latimore analyst
#16

Great, just one more thing. Would you walk us through how the debt is being financed for Indonesia? A handful of quick questions. What's the interest rate — does the customer cover part of it? If yes, what amount? And what share of the financings has been finished in that location?

Jayesh Chandan executive
#17

There's an offer we put out in the press release regarding this project covering 70% of it. The remainder is going to be funded by the grower's balance sheet, plus customer prepayments. So the way it's worked up to now: when we deploy 30% of the project upfront, or 30,000 servers, that portion is paid for through customer prepayments and through < >. After that, the debt piece would cover the 700 servers arriving at the end of the year. That's everything I'm able to share regarding how this project is being financed. So the earliest deliveries are essentially the clearest pocket, and then they join on customer < >.

John Roy analyst
#18

Nice work, everyone. I'd like to pull back for a moment and consider what might go sideways in the coming year. In your view, what's the largest risk you face? Could it be staffing? Could it be electricity? Could it be putting up buildings? Could it be getting hold of machinery? What keeps you up at night the most?

Jayesh Chandan executive
#19

Those are excellent questions, John. It's good to hear from you once more. Let me break this down into 3 main categories of risk, all right? First comes hardware timing. Second — I'd put this as site and power readiness — and third is customer acceptance along with workload migration. We've been placing orders ahead of time — just so you know, every piece of equipment that was meant to arrive at the end of September has now been delivered. That's roughly 5 weeks ahead of schedule. Yotta 2 comes in around 8 weeks early where manufacturing and similar matters are concerned. So we order ahead, we test prior to deployment, and we accept deliveries alongside our OEMs and integration partners. Then, even more crucially, we ensure that sequencing is properly handled. Next, we were gating deployment on what I'd call confirmed power and ready site readiness. Engineering is hard at work 24/7 on this networking effort, and installation crews are operating simultaneously. Our people are stationed at sites across various regions of the globe; sadly, electricity has an awkward tendency to be absolutely necessary. On top of that, customer testing plus workload migration must start prior to full commissioning. That let us work through every integration problem step by step. What we're doing, then, is ensuring the paying workload performs better. So we've built scheduled buffers into each of these areas and base deployment plans. And even more crucially, we have the capacity to resequence work whenever it makes sense. We also need to establish our own internal processes. You'll notice we've actually gone — we've nearly doubled our headcount on the human resources side; and we're also ensuring contingency plans exist, covering both operational preparation and all the hardware too. So every one of the < >, every RFS, all of it must be finished well ahead of schedule; that's something we consider among the key risks we're watching during the H2 ramp for ourselves. Still, we hold contingencies for each and every one of them. John?

Bharath Nagaraj analyst
#20

Earlier, you brought up aggressive hiring — that's the term I'd use — throughout Thailand, India, Taiwan, and so on, aiming at over 1,000 individuals, if I recall. Possibly just Thailand by itself. Where does headcount stand right now? And what does the monthly OpEx run rate look like as we leave Q2 at this point? That's my first question.

Jayesh Chandan executive
#21

< > you once more. So on a full-time resource basis, I believe we're sitting at roughly 300 < > and on a contracted basis, we've already reached somewhere around 300, 350. As you can well imagine, we're moving through a number of phases. At this moment we're working through land assessment, power readiness, water readiness, EPC certifications, and so forth, in both Thailand and Indonesia. That will likely bring in another 400 to 500 people, though a great deal of process is underway right now as we speak. Looking ahead, we'd expect that figure in particular to grow substantially, so I can share that at the peak — say, by mid-2027 — we'll have roughly 2,000 people. Right now, we're at about 300 plus another 50, so around 650 people; so indeed, we're growing quickly. Bruce, would you like to handle the second part?

Bruce Bower executive
#22

Yeah, the SG&A runs roughly $2.7 million per month. So it covers more than just personnel expenses. We anticipate that figure will climb by the close of next year, reaching an annualized range of about $38 million to $40 million. As Jay pointed out, we're adding a lot of head count, but bear in mind that our locations in India, Thailand, and so on aren't Silicon Valley, correct? That's the reason the cost bump wouldn't be all that big. And we believe that's — to run projects bringing in hundreds of millions of gross margin, I'd say that's a trade-off worth making.

Bharath Nagaraj analyst
#23

Got it. My next question concerns GPU spot prices, since they swing around so much. Do Gorillas have take-or-pay contracts at a set price? Or does it track the spot market — say supply catches up next year, and you're locked into a multiyear deal — is Gorilla then carrying the margin risk? How does that actually work?

Jayesh Chandan executive
#24

Great question. Essentially, every piece of our hardware carries a fixed cost at present. Nothing operates on what I'd call indexed or pass-through pricing, which would shift some of that volatility onto their customers. Every one of our agreements — the power agreements, for instance — is fixed. Power prices don't move, and the rates we pay for water and connectivity are locked in too. So we aren't attempting to hand the risk to customers, nor to hold onto it ourselves. It's all been spelled out very plainly. Our customers have been extremely understanding as well. That means a bigger charge upfront, and essentially we say: look, here's the risk right now. Here's what memory costs; it climbed 40%, 50%, 60% across the past 4 or 5 months. This is the updated cost, and we're making certain we intend to safeguard the economics of every project — we're not simply leaning bit by bit on unhedged spot pricing or anything of that sort; that being said, we've visited each and every vendor in person — we haven't sat down with all of them and confirmed the prices are fixed. So anything they deliver from now through December sits on a fixed, fixed-price basis, and it carries no risk for us.

Bharath Nagaraj analyst
#25

All right, that's quite straightforward. Just a handful more items from my end. No problem. One brief point. I see you gave attribution. Project-level guidance for 2027. What I want to understand is what utilization assumptions underlie that guidance, maybe for '26 and '27 as well? Then there's a separate matter regarding Egypt, Taiwan and Thailand. What kind of contribution came from them in H1?

Jayesh Chandan executive
#26

No, good question. Bharath, I'll break this into two pieces; there's no project level guidance on attribution — use attribution, because these are take-or-pay, all right? Whether the customer uses it is up to them. It hinges on an identified customer, not on customer demand, all right? So in our case the mix is commission GPU capacity utilization at 100%, that's what we assume. Of course, some RMA issues exist. So the customer expects a 99.95% delivery. But more importantly, we're making sure every one of these recurring compute service revenues belongs to the established business. So when I say I'm billing, just as an example, one for this customer, that won't fall to 0.9. It won't rise to 10 — it's actually $1, and we'll be billing that over the next 5 years. Now we model a face commission. The commissioning is the portion we'll share. I mean there'll be an onboarding customer per as utilization rises as workloads migrate. So here's what happens: customer testing occurs, the first 300 servers go live, the next 700 go live, then the next < >; that's the onboarding curve we have — though regarding payment, the customer actually pays us a flat fee for it; now Yotta clearly is a big contributor as the phases progress. India and Indonesia are set to become very aggressive. It'll form a huge part of our revenues going forward. Yet at the same time, on Egypt, Taiwan and Thailand, we've already had, as Bruce alluded to, we've already established a security, network intelligence, public sector operations and so forth; we're not abandoning that. In fact, we're bidding on some very large projects right now, and we're going through the motions of closing them as and when. So we'll make the necessary announcement there. For us they remain a very important source of revenue, customer collections, but more important technical credibility to help support our AI infrastructure. So let me explain, for those who aren't clear on exactly what kind of company we are and how the mechanics work. Everything — our network intelligence, the security systems we've put in place, video, the platform handling data intelligence, building management offerings, and the rest of it — sits inside our data center, our stock, and the NOC, and that NOC, which we constructed for Egypt, has genuinely turned into a piece of our business. These days we're delivering that same capability for our own data centers. Alongside that, we put money into a Bangalore, India company named < >. Their role is to assist us in combining our inventory alongside our BMS offerings, and they provide a more robust tech framework that we depend on ourselves at certain data centers. Consider < > as an example: once we have it constructed, roughly 90% of the technology incorporated will come via Gorilla or through < > white-label products, which ties into the $0.9 figure we discussed.

Bharath Nagaraj analyst
#27

That's quite useful. Those details help a lot. I truly valued hearing that. Many thanks.

Jayesh Chandan executive
#28

Thanks a lot. I mean, thank you — gratitude to all of you, for taking part in this. My appreciation goes out to each person who stayed with Gorilla back when the figures were far smaller, when our story might have needed a map, when folks questioned certain nations, when there were moments of doubt and honestly a decent dose of danger. Over the coming 12 months — well, look at what we've accomplished since then: we've doubled revenue for our first half. Our operating loss has come down by roughly 95% versus Q1. I'd love to claim that as my own doing, but sadly the individuals responsible — the ones who truly put in the effort — are in fact on this call right now. So credit isn't mine to take. Now my < > posed a question to me. She asked, hey, what's on your mind? My answer was power distribution. For my part, I'd say romance isn't dead. All it takes is a substation. That's precisely what we're constructing. So in my role as CEO, I supply the optimism, while Bruce calmly points out that, hey, under the IFRS that optimism still isn't acknowledged. < >, regrettably, he doesn't flinch. So I'm requesting that each of you assess us based on the contract you put your name to, the equipment we've shipped, the revenue we've booked, and the cash we're gathering — for every shareholder, every customer, every partner, and every one of our remarkable staff members, both all the fresh ones, truly, thank you. Our goal is to make your bets look not so much like safety and much more like brilliant choices. I realize at times the market can weigh in and they'll say, you know what, oh my goodness, the CEO was out to make a fool of you. That's alright. No problem. Gorilla has barely begun. Say whatever you like. And once everything unfolds as planned — and I'm stating this plainly — once everything unfolds as planned, not these. Eventually my family will welcome me back to the dinner table, the one I walked away from, I give you my word, we'll be discussing GPUs and megawatts at the meal. So until that time, folks, thank you so much for spending your time with us, and thank you for standing behind us. Cheers.

Operator operator
#29

Everyone, that wraps up the call for today. We appreciate you joining us, and you're free to hang up at this point.

Read the full transcript via the API

You're viewing the first half of this call. Get the complete Gorilla Technology Group Inc. transcript - plus 132,000+ transcripts from 7,000+ companies, speaker segments and full-text search - through the EarningsAPI REST API or hosted MCP server.

Get an API key View API docs →

For developers and AI pipelines

Programmatic access to Gorilla Technology Group Inc. earnings transcripts and 132,000+ others is available through the EarningsAPI REST API and the hosted MCP server. Quarterly plans from $162 - full transcripts, speaker segments, full-text search, and the /api/v1/transcripts/recent polling endpoint for ETL pipelines.