Perfect. I think maybe we can start the show, remove the banner. Welcome to this presentation with the Lokotech Group. Today I have with me the management from Lokotech. I have CEO, CFO, and Chairman. As always, if you want to ask questions during the presentation, during the live presentation, you have to go to Lokotech's profile at Investorweb in order to submit questions.
So if you are watching this presentation at Finansavisen quarter or social media, you have to go to Lokotech Group's profile at Investorweb in order to submit questions. The format for today is short slide presentation. After the slide presentation, we will have Q&A. We got an extensive list of pre-submitted questions, and we will also try to answer questions that you submit during the webcast. With that, I give the floor to you guys.
Thank you, Ubaid. Thank you for the introductions. My name is Ola Stene-Johansen. Today we have Yngve Johansen with us, our Chairman of the Board, and my Co-Founder and CFO Benjamin Miklozek. This webcast is being hosted because of the latest upturn in the cryptocurrency market, and we're trying to encompass and answer all the shareholders at the same time. So, a little bit different than a fireside chat. But still, we have a slide deck. Some of those slides were meant for pre-reading. I'm not going to spend too much time on the deck itself. We'll briefly go through it and try and spend as much time as possible on the Q&A part.
First, a Safe Harbor Statement. To introduce the case, I'm sure that a lot of the viewers are familiar with us as a company and the history we've had since our IPO in 2021. Back then it was a relatively pure play ASIC investment case. Over time, since then, we have developed some other interesting verticals, as you may know, both the data center business, PowerPool acquisition, and some smaller verticals that are interesting to talk about.
I'm going to spend too much time on Filecoin today, try and focus on the ASIC and PowerPool and take it from there. Lokotech Group is a fabless semiconductor company, meaning that we don't own the production facility ourselves. That would place us in the multi-billion-dollar market cap category. We've been working and developing our dual-use ASIC since 2019, and the R&D phase is now finished. We are in the process of moving into the next phase, ramp up for production.
We have had some interesting news come out lately, both with regards to PowerPool and the growth that has been supporting a highly profitable entity that we are optimistic about in the future. And also, we've gotten some more meat on the bone on the edge AI part. So, that being said, this is our group of companies. I'm not going to spend too much time on the history. When we received the final power simulation that marked the end of the R&D phase, in a sense, at least the research part of it, we are highly confident that this is the final specification of a commercial ASIC, meaning that the accuracy of the simulation tools used to get to this data has a 5% accuracy up and down.
Final power consumption of the final unit will also depend on local factors like the voltage on the grid, temperature that it's running in, and so forth, which is why we, as any other manufacturer in this space, operate with an interval of 10% ± . We have included some quality features that separate us from the competition. And right now we are at the TRL 5 level when the engineering wafers, that means the first wafers that are being produced on the commercial silicon are being processed, that moves us into the TRL 6 level. And from there we are building the value chain to quickly move into TRL 8. It's a full-scale tape-out.
If the yield curve on the production wafers are good enough, we consider it to be TRL 8. So, as you know, power consumption is the main operational cost of miners today. Well, with increased profitability and lack of hardware, you could probably argue that overhead is now starting to kick in as a large contributor. But regardless, for long-term viability, long-term profitability, power consumption is key. And we use 35%-80% less energy than the competitors to mine the same amount of coin. So that's a key differentiator on top of all the quality of our features.
So, the ASIC is the engine, right? And we'll be around 90% of the bill of materials for the product, regardless if it's the newly released Single Barrel and Double Barrel units, or if it's the Hashblade. So, the form factor is just a matter of suiting the mining hardware to each client's type specific needs. We have previously focused on marketing the Hashblade, trying to reach a wide audience, ease of use. We turned it more towards the industrial-sized miners on the latest release of the Single Barrel and Double Barrel.
And as we have discussed earlier, we can move up or down the vertical, use the equipment ourselves to mine or sell the chips to OEM manufacturers already in the space. So, a little bit important to understand that, it's not necessarily only these products that can drive the revenues in the future. I assume that you know us. I'm not going to spend time on the CVs of these gentlemen. But I think the key takeaway from a technical standpoint is the fact that we are using a 12 nm low power plus production node. It's both cheaper to design on, cheaper to produce on, and it's competitive with regards to accessing production capacity. And it also future-proofs the design if we decide to take it to a lower node and more advanced production node, like, for instance, the 5 nm process that our competitors use.
In addition to that, they're going to be opening a line, I believe, in the middle of next year for a 12 nm in Germany as well, which can accelerate our wafer production and free up even more line space.
That's a good point. GlobalFoundries has been investing in their production capabilities in Germany. So, we've always said in New York, strategically we're keeping the value chain outside of Asia, but it's good to know that GlobalFoundries is investing into the future of this production node, as you mentioned. So we are still on target for the scheduled tape-out date.
We have booked a date with GlobalFoundries for when the tape-out will happen, and it depends a bit on your definition on the tape-out process, because what happens is that we conduct what's called a Design Rule Checking internally to make sure that, according to GlobalFoundries' rules, we can produce it. 99% of the chip's area has already been concluded, and is DRC- clean, so there's very little job left to do on making sure the rest of the chip is also DRC- clean.
Then we hand over the design to the producer. The producer conducts the same DRC to make sure that they are able to produce it. If there's any risks, we are able to sign off on waivers, meaning that we take on the risk rather than the producer. This is a process that every chip goes through. Global is not interested in spending millions on mask sets and wafers without actually knowing that the chip is producible. Last time we conducted this DRC took us a week and a half, and it was considered the most clean DRC design they've seen. We are scheduling a month's worth of time for that process. When that mask is ready and done, we will then initiate some shuttles of engineering wafers, high risk, no responsibility from the producer on things like yield.
And those units will then be what takes us from TRL 5 into TRL 6. They can also be used as customer samples, full-scale prototypes, or we can put them to use ourselves. Eventually we can sell them. It all depends on the yield and number of chips we get out each wafer. When those engineering wafers have been completed, we are then into a production wafer situation where we order chips and they come off the fab and, well, mount them, sell them. Anything to add?
I mean, other than the timeframe for actually increasing yield, I mean, everybody, every company out there that brings a new chip to market always has low yield as the manufacturer optimizes the tooling to maximize yield, and just for numbers' sake, that usually happens between 800 to 1,000 wafers, then your yield is maximized.
So when I warn investors that there might happen a Revision 1.1, that could either be because we want to fix something that we don't see today, or more importantly, we want to optimize the economies of production by increasing the yield by making certain adjustments to the mask set. It's also something everyone goes through, and it's a decision that we have to postpone until the first production wafers have been completed.
As you know, we have announced and have voiced ourselves in annual reports and half-year reports for some years now that we do see a lack of financing to initiate mass production. That is, of course, our number one priority. Specified now a little bit more in detail what is required. But I want to point out that the number of wafers and capex can be adjusted up or down, and it's not firm in that sense. We can also raise more to start more production wafers if needed to be.
And as long as we are confident after the first engineering wafers that we are able to deliver the actual products, we can start using the client's funds to fulfill those wafer orders. So at some point, we will move into negative working capital territory, and that will mark significant milestone, at least business-wise for us. The competitive landscape hasn't changed much since our last presentation. Bitmain did launch the L9. We're starting to see clients putting online now, a bit delayed. ElphaPex was a new entry that had a good, what is it, four, five months to take a chunk of Bitmain's market share?
Yeah.
But we haven't seen any indications of their new generation or next product line from ElphaPex. I was challenged to try and update our sales expectations. And first of all, I think it's an important topic to discuss this number of visible hash rate today. It equals about 700,000 hash rates, right? So that's the input for this array. Whereas in reality, the prices of the underlying mined cryptocurrency is what determines how much hash rate the networks can sustain over time. So we might be underestimating the total at the attainable market, but we had to start somewhere. And I think replacing the hash rate or doubling the hash rate or as equal is a good place to start. We do know that the market has, you know, winner takes it all attributes.
The more efficient you are, you will now compete with your older generation hardware. Right now, though, with the current rising prices, everything is profitable. This is to consider on a little bit longer timeline. We have access to pool data now after owning, or acquiring PowerPool. So that makes our deterministic predictions a bit more precise when it comes to the expected economic lifetime of the units. So we are now telling you, for selling the units, we believe we have three years' worth of time where it's interesting to buy. The total economic lifetime of operating the units with historical difficulty increases, taken into account, we're looking at seven and a half years.
With that, you got to take that with a grain of salt, because we're currently running the L3, which came out way back in the day. It's profitable.
Yeah, it' s profitable. So, there's a gap in the market when it comes to Scrypt mining hardware.
But the major benefit of this, and miners do realize this, is it's not always a bull market. I mean, it'll always retract and go into a bear, but that's what's good. And that's what we're going to lean into is you can run during the good time and you can run during the bad time. And that helps the, you know, the economics behind the, you know, the small, mid-sized miners that don't have the capabilities of, you know, complete optimization to getting power for, you know, near zero prices. So that helps out with decentralization and helps out with our client base, as we move forward.
With this slide, we want to illustrate two things. Number one, Dogecoin has become a really important metric to follow. That's a little bit of different dynamics than Litecoin, where there's an infinite supply, whereas Litecoin has a restricted supply of 84 million coins. So Dogecoin has become an important, important part of the profitability of mining Scrypt. That being said, we also saw Bellsc oin being launched as a merge-mined coin with the two others. And it builds the flexibility of the Scrypt algorithm compared to, say, SHA-256. We expect more coins to be added onto that merge mine list. Ingenious invention and a sudden increase in profitability for Scrypt mining.
Yeah, it's a, that's a huge benefit to have more daily emissions come onto the same chain as Litecoin. And this is, you know, at the blockchain level. I mean, in PowerPool, it's good to mention that we mine up to eight to 15 coins at once, merge mining with them, depending on what's profitable. You know, to name a few, you got Luckycoin, DigiByte, Pepe. Soon this next mid or not, maybe in two months, we'll have Meowcoin. So it is, and that's, you know, a strategic advantage for at the pool level to squeeze out as much revenue as we can, with PowerPool.
And ASICs.
Yeah, and ASICs.
I want to talk a little bit about PowerPool. The pie chart here is for SHA-256. Bitcoin, Bitcoin Cash, Bitcoin SV. Actually, I promised myself never to mention that name again, but regardless, the key takeaway is that within the SHA niche, PowerPool is still a very small actor. It's growing rapidly, but there's a lot of market share to be taken. Some of those growth initiatives are designed specifically to take more market share from institutional miners in the U.S. We also included a year-to-date benchmark. The date here is early November.
Yeah.
Just to show that, we're not only profitable on a 24-hour basis, we're also profitable over time. And I am of the firm belief that capital in this sense, the miners, machines, and hardware will be attracted to the most efficient use. We did release some record numbers, and we felt that it was a good idea to also include some averages. And when you're talking about pool business, the gross turnover makes great sense. It can be very large, and you can lose money off of a multimillion-dollar gross turnover, because it's the net fee that really marks your income. So for November, at least 1st until the 27th, we're averaging EUR 3,400 in net revenue. And I think it illustrates the explosiveness of that business plan. It has grown around 14% compounded monthly during the last year.
If you take that growth rate into as an assumption for 2025, that would equal around, sorry, EUR 2.8 million in net revenue for PowerPool, the entity. If you are a bit more conservative, we're operating with three months, sorry, 3% compounded monthly growth rate. Then the expected net revenue, or actually profits before taxes, is more or less the same, in the range of EUR 1.5 million . For the AI use case, we've made some discoveries. We also started to benchmark it at least on a theoretical level against the competing edge AI inferencing ASICs out there. We tend to beat the incumbents either on computational power per chip, computational power per watt, or computational power per dollar.
So, if the simulations are correct, and I want to remind you that this has not been tested on a full scale in the commercial ASIC yet, it still is on a TRL level 3. But if it's correct, we have a lot of different product market fits that I think will become very relevant. And over time, that could be a more important revenue stream for us than the Scrypt mining vertical. So that being said, even though the TRL level is relatively low, it's the same chip, the same chip design. So we're going to jump some of those latter steps and go straight into TRL 7, 8, when the tape-out has happened, the first engineering runs are done.
Then just to add on that, as you know, we come to fruition of testing for the AI and everything works as simulated. What's going to be really good is not only setting a standard on the legacy node, as we've mentioned many times, but we're getting a dual benefit by, you know, taking the design down to a lower node, which usually takes about nine to 12 months to hand it over to begin the masking process as the position we're at now. But then what that equals down to for both the Scrypt and the AI is basically lower real estate for chips on the board. So then you can put it into, you know, smaller products instead of having the chip the size we have on the 12 nm, which helps out a lot.
Sure, sure. Another interesting part of this AI business case is that to our knowledge, we'll be the only ASIC that's made for inferencing, but also have the capabilities of handle model training and at least feeding back to the cloud, input for improving the model. And that opens up some very interesting niches, like the gateway market, where internet gateways haven't really had any major development since the, what, late 80s.
Oh, long time.
Yeah. The right for replacement and having nodes or AI inferencing capabilities in the future internet, I think is a very interesting prospect.
So our burn rate is still around $225,000 quarterly. It hasn't changed much since the IPO. What has changed, though, is that the free cash flow from PowerPool Mining, at least if you take the minority interest out of the equation, should make us cash flow positive the full year of 2025.
It's obviously dependent on prices of underlying and also the growth rate that you put into the assumption. As you know, the B2C pre-sales have been slow. These figures are not including the latest release of the Single and Double Barrel. But I'm very happy when we announced that we finally got the signature for our framework agreement. It is a relatively flexible agreement. The client has indicated that the interest to be in the upper range of that, for that interval. We are continuing dialogues for different structures of agreements like this or less flexible ones, so to say. It is true that the second deposit, the final balance, will be due before production so that we are able to enter that negative working capital case.
For those who read the annual report, I just want to mention that we have a relatively strict amortization of the R&D investment, and that all intra-group debt has been written off and will continue to be written off. We have no debt as of today, but we do have a seller's credit for the acquisition of PowerPool. So we, we're a fabless semiconductor company. We have other interesting verticals, but that's the main, main vertical for us.
We are expected to deliver the first units in Q3 next year, and the chip has use cases into multi-billion dollar high growth verticals, so that's what I have for today. Ubaid, I suppose you have some questions.
Yes, guys, I have some questions, but we can maybe, Yngve, you can start with introducing yourself.
Yes. So my name is Yngve Johansen. I'm a physicist of background from NTNU, and I worked with technology for a long time. I'm coming from the oil and gas industry, but also worked on the lab. Being very interested in the blockchain industry, I also joined in the Board to Lokotech and now I'm the Chairman. Yeah, and I'm just an enthusiast on technology.
Perfect. Thank you. With that, I just wanted to let you know that we have received your question. Okay, guys, I think we can start with the pre-submitted questions. Once relevant, we will also read out the question in the chat here. We have categorized the questions. We will start by talking a little bit about the framework agreements, orders, and commercial terms. Is that okay?
Sure. Do you want to answer it, Ben, or?
Yeah, I can answer that .
I can read out the questions, and you can answer them.
Okay.
Yeah. Could you elaborate on terms of the recently announced framework agreement, specifically what is required for the second payment and final order value to be confirmed?
The second payment will basically confirm the final order value. And as Ola was talking about earlier, you know, that would be due before the production starts so we can take advantage of that capital. As for the terms of the agreement, I mean, we really, this is more for the miner side and why it is very welcomed in the industry after, you know, we've only shown it to a few people after we've kind of restructured it.
But, you know, having like 365-day manufacturer warranty, having a return policy for the advanced RMA, kind of adding more westernized customs when it comes to, you know, having confidence in the product that you deliver, you know, really helps transform, you know, those people sitting on the sideline and wanting to order, I feel it is. That's, you know, one of the bigger reasons why we got this order for the two to five. And that was actually for a Hashblade order, not the Single or Double .
Great. Thank you. Which factors will determine whether the final order value ends up at the lower or higher end of the estimated range?
So it'll be determined by the balance settlement. So far, the client has put up a security deposit for the full five. And when we have the specific timelines on delivery for him, he will decide if he will fill the whole thing and take five or if he will only take two.
I can also interject on that, Ubaid, that, of course, it's the customer that decides how much he wants to go with. It's not like we are going to tell them about that, but the initial indication is $5 million. Yeah.
Okay. Thank you. Some questions regarding tape-out and technical progress. You mentioned that the tape-out must be completed before final order confirmation. Could you explain why this is so critical and what milestone we can expect to see during this process?
Sure. So, we have the tape-out date. That's the target everyone is working towards. There is little engineering work to be done other than the DRC of what's not already completed. So we are very confident that we'll meet that timeline. But towards the client, that meeting that deadline will be important of determining when we can assume we can deliver the final product, if that makes sense. I think it's also important to note that this tape-out date is not like the MPW dates that we have been chasing after earlier. This date is flexible and can be moved back and forth depending on the situation then and there.
Sorry, I have a follow-up question regarding this. Can you clarify what remains to be done to reach this goal?
Sure. So what the engineers are doing right now is, they're doing a process called LVS and DRC. That means that they have a digital design, and they want to make sure that the digital design is also represented correctly in the physical graphical design, layout versus schematic. So, if there's anything that they discover needs to be improved or needs to change, they can do so, manually or with an auto router for the digital part of the design file. So, 99% of the area of the chip has already been cleared for DRC- clean, which means that we hope that the DRC process will take a shorter amount of time than what we are scheduling for.
Yeah. Thank you.
I can also interject on just, on that one. How long will the DRC process take at the foundries? And it's one of the reasons it can be. We have, you mentioned earlier, Ola, we have taken into account that it will perhaps take four weeks, while earlier to the MPW it took one and a half weeks. So we hope that will happen again. So there will be, you know, there is a certain amount of movement in the actual start of creating the mask set, right? So that's why also it's, I saw there was one question, why can't the exact date be set? It relies on many moving parts. Some of them might take slightly longer, some might take slightly shorter.
To add on to that, Yngve, we have no indications as of now that we will miss the date early Q1 next year.
Thank you. Regarding competitive differentiation and market standing, Lokotech claims to have a state-of-the-art Scrypt ASIC. Could you detail which features and performance metrics make Lokotech's solution unique and attractive, especially compared to major competitors like Bitmain?
Sure. Do you want to use sales pitch, Ben, or what?
I mean, we are just a lot more energy efficient. I mean, if you're taking off a percentage we're beating Bitmain, who's on the 5 nm that we suspect, by 135%-140%, just in power consumption for hash rate. And that's a huge advantage, as I mentioned earlier, for the bear market. And because when the bear market goes down, people start hunching down and trying to just make sure that their facility can stay online. Well, our machines will be the last ones to go off because as we start to add hash rate, our Hashblades, Single Barrel, Double Barrel start taking more and more daily emissions from these legacy miners, whether it be from ElphaPex and/or Bitmain.
And then, you know, those daily emissions are coming to us while they're losing, and it's more tighter. Unless they have free power, they won't be able to stay online. So as those legacy machines fall off, our Hashblades and the more efficient machines will grab those daily emissions. So it'll just be a hierarchy as we kind of keep putting hash rate out to market. That's the number one advantage.
As we mentioned multiple times during this presentation, since we can take this down to a node within 9- 12 months and then go straight into masking and production, it really protects us from any competitor coming in with more efficient ASIC. Because if we go down the node, you can basically double the hash rate and half the power consumption with pretty much the exact same design. Now the competition, if they're bringing a new ASIC to market, they just won't be able to compete as we keep delivering hash rate to, you know, B2C, B2B.
On top of that, we have the gradual degradation, meaning that if a part of the chip breaks down, you won't have a critical error or a failure. The microcontroller on board the chip will turn off that engine, and the rest of the chip will function as normally. So, instead of having a massive degradation happening from one chip failure, it is gradual over time.
Then even to add on to this, I think that we'll probably see more manufacturers, but we're going to be the first manufacturer that has a front-end PSU that basically slims down the footprint of your miner, being able to stack more units. You know, PSUs, they fail a lot, and the quality is not that great. But having a server-grade front-end PSU is cheaper, it's more reliable, it's higher quality, and they're available everywhere. So this way, you wouldn't have to order it from, you know, China or order it from some other area of the world that's a third-party manufacturer. You'd just be able to get it, whichever country that you're in, given the amount of data centers and HPC, AI, all that stuff that's using them.
I think the efficiency, there's also another argument we had, and that is your infrastructure cost. So you invest in, say, 1 MW worth of capacity. Now we don't need to add another megawatt. You can, you can get a higher revenue per megawatt by being more efficient with the use of the power. So your total hardware investment will then increase relative to your infrastructure investment. Combine that with a smaller footprint, you have a pretty good advantage.
Yeah. In addition, there's, when it comes down to the software as well, we are implementing a bunch of stuff, even like, you know, moving to a USB open protocol for the communication of the miner helps out a lot as well, because it really does make it more of a plug and play than what the typical miner is nowadays.
For my sake, I'm just most enthusiastic about the fact that, you know, blockchains are supposed to be, well, not necessarily all, but many people like blockchains, decentralized blockchains, because it's something that is run objectively by itself, not in a few units' hands, and if you have power-efficient units like what Lokotech is producing, soon you basically can have units running out of cities. It doesn't have to be in the middle of the desert of Texas or some other place with really low power prices. It can be in anybody's house, anybody's home, because it will run with profit when it is so efficient. That's what I really like about it.
Yeah.
Thank you very much. Could you compare the Double Barrel and Single Barrel Scrypt miners specifically with the competitor's product in terms of hash rate, power consumption, and price?
It really comes down to, I mean, our price right now is, you know, beating them. I haven't seen the latest, future pricing for, May deliveries, but, we're beating them on, on all aspects. Given that we should, we're able to get a premium for that, power efficiency is quite huge. I mean, from, from our standing, from a competitive advantage, when we launched the Single Barrel and Double Barrel, ElphaPex put out a, pre-order for, May deliveries, 45 minutes later, and then Bitmain put out deliveries 30 minutes later. We know that they're watching us, and it's, it's going to be fun. I love competition, and it's, once we start delivering, it's, it's going to be fun in this winner-take-all market.
To be precise now, it's higher hash rate, lower power consumption, and better price.
Yeah.
Thank you. Market reception and pipeline. How large do you believe the total addressable market is for your ASIC and miner products over the next two to five years?
So we touched on it briefly in the presentation. We are operating internally with doubling or replacing the current hash rate as a target for the TAM. In reality, that market could be a lot larger, at least if the prices of the underlying are at the current levels. But I think it's ambitious enough when you're talking billion dollar plus in revenue figures. So there's no reason to doubt that there's a large and significant market out there that we can attack. Then you end up in an assumption of how much of that market can you really conquer. And as I've tried to explain many times, it's a winner takes it all market over time. So unless the competitors have something up their sleeves that we don't know of, it should be good for a long time.
Yeah. Could you talk a little bit about which customer segments you see the greater potential in?
So that's a difficult question to answer because, politically speaking, we are, we have and we want to keep the minimum order quantity at one blade, meaning that a teenager in theory can start mining. Whereas for us, business-wise, it's obviously more interesting to focus on the larger accounts. But as Yngve said, anyone can run it, so there's no reason not to offer it.
So here's a question about long-term strategy. What are the company's long-term ambitions beyond 2026? Do you plan to develop ASICs for other algorithms or applications, in the future? And how far along are you in these plans?
So I can speak on behalf of myself. I can't speak on behalf of the company because the company hasn't focused on panning out the long-term strategy. It's been focused on overcoming the near-term challenges. But for me personally, I look to the group to become an important player, not only in the mining industry, but also in the AI sector. That being said, we do have some ideas of how we should go after the SHA-256 algorithm. We do see that the current design, which is a Scrypt, is a memory-intensive algorithm. So it's natural to think that a lot of the efficiency gains that we have had is also applicable for other memory-intensive algorithms, like not SHA-256, but Ethash for instance. So attacking other algorithms, and of course porting the current design into a lower node, are priorities from that vertical.
The edge AI space, I think that, licensing this design or the next design to an incoming player that has the distribution network, that has the capabilities to manpower the, the finances to really take a large chunk of the AI market, it's going to be key. So then the question is, is it a JV or is it a sale or too early to determine? But, if you take Lanner, for instance, that operates in this market, they have 1,200 employees on all continents, and we're not going to grow into that type of a setup overnight. So that's my ambitions anyways.
Thank you. You mentioned having fixed markups in your agreements, but could you provide more details on how margins might develop with larger volumes and how you achieve economies of scale?
Sure. I also mentioned earlier that around 90% of the bill of material, meaning the final cost of the whole actual product, is related to the ASICs. That's wafer cost, and Ben mentioned that you start to see the yield and also the economies of scale when you reach a certain milestone worth of number of wafers, so economies of scale and the margin will then depend on the volume that we are able to sell or put into use ourselves, and that's why I can't predict accurately which part of that 30%-50% gross margin range we will end up at, but this agreement is based on a direct cost plus a fixed margin, meaning that if the client waits until we have achieved relatively higher volumes, he'll partake in benefiting off of those economies of scale.
If your technology proves to be market-leading, how would you scale up production given that you already, I think you answered it, but we can read it again. If your technology proves to be market-leading, how will you scale up production given that you already see interest from multiple, potential customers?
Well, it comes down to the sales volumes, right? How many do you sell or how many do you capitalize to produce for your own use? So, it's a volume game in the end.
Yeah, I can also interject. I mean, it's you have to remember that we also, if you are just making any chip, then you have to rely on selling it to somebody and it being popular and gaining market and all this. So, of course, we want to do the same thing, but due to our multiple verticals, we are also in a position where we actually can run it ourselves. So it gives us, I feel always there is, for us, it gives us somehow less risk in that regard. We have some options when we are set up the way we are. I don't know, don't you agree to that, Ola or Ben?
Yeah, I mean, that's a huge benefit and why we've always wanted to be vertically integrated all the way to the chip level, because in this industry, there is no warehouse risk. If somebody doesn't want to buy it, you plug it in. It's, if somebody's mad at you, you plug it in. It's a very easy industry compared to anybody else and to not have that warehouse risk, because, I mean, being in other companies, I know that can be stressful when you're trying to balance the, you know, supply and demand when we are our own demand if we produce and/or sell.
More importantly, I think, is not only the actual sale, but the price. Sale. So it's a matter of, is it more beneficial for us to sell it or use it?
Because we'll have the quickest payback out of any miner for when it comes to plugging in ourselves. And then also for the reaching economy of scale, I'd like to mention this is extremely beneficial for the AI aspect of our ASIC, because yeah, you know, companies can produce a lot of AIs, but some, you know, AI chips and, you know, go after that vertical that they only have. But since we are a Scrypt miner as well, we can, you know, confidently suspect that we're going to be ordering a ton of wafers, which will, you know, allow us to get the economy scale faster. And then that's also going to be passing on, you know, that economy of scale to the AI as well. We're offering cheaper chips than what the customers will be able to as well.
Competition.
The competitions.
Yeah. And we are also talking to other people in the industry, in the silicon industry, making designs for wonderful chips and, but they are basically dependent of that chip being popular. Sometimes you might even have a great product, but for some reason, it didn't sell at scale. I always felt that that risk is not very for us because we, everything that will happen, we will use it ourselves.
It also empowers the RMA because if the client is unhappy, hash rate is lower than anticipated because a chip or two is bad. Okay, we replace it, we run it ourselves. The efficiency is the same, it's just a matter of hash rate per unit. It empowers that type of Western thinking on customer service.
Thank you very much. Now we will talk a little bit about finance, financing, capital market needs and financial projections. You previously indicated that you may need additional capital for mass production. How are you considering the various financing options?
No, so I think that in the timeframe that we are currently in, the pre-order type of financing is probably not going to happen right between Christmas and we are taping out early Q1, so unless Ben is able to close a really big sale, that's a little bit off the table at this point in time. That being said, and we also discussed this in the website fireside chat in late October, we are in continuous dialogue with different ways of financing it, from debt to hybrid to equity, so it's not firm yet. It is the number one priority, at least between the upper management. That being said, upper management, they're also large shareholders, so we will opt in for the option that will benefit our shareholders the most.
Thank you. I saw there was also multiple questions on that. And one other thing we could add is like, for the investor, it would be sort of useful if they can look on the mask investment, plus the silicon investment, because, you know, the amount of wafers, the amount of silicon we produce, that can vary. We can be many different scenarios also. So, when you're looking on the numbers, it was one of the slides that Ola had earlier. It's useful to sort of divide it up like that, I think, to understand the needs and the possible scenarios also on the production investment side.
Thank you. Carlsquare projections and guidance for 2025. Carlsquare projects Lokotech sales revenue at approximately NOK 9.1 million , aligning with figures from your H1 report for 2026, they forecast NOK 53.8 million, roughly matching the value of the framework agreement. Can you clarify whether Carlsquare derived this estimation independently or if they're based on guidance or input from Lokotech?
I can comment that, when I was, I would say lead, because we only had one meeting with Carlsquare just to try and bring him up to speed on all the different verticals that we are operating in. Any estimations and projections they have made either out of publicly available information about us or their own predictions of the market. So we haven't given any input to them other than what can be read in an annual report or presentation that we previously had.
That being said, the Carlsquare is considered the best or one of the best analysis providers in Germany. They came up with a target price of 2.1 on the initial coverage report. They have a really good track record on hitting that target price, but I see it as my mission to break their record and put their estimates way off for the next time they're trying to benchmark themselves.
Thank you. Yngve, for you or you guys mentioned that a major industry player has shown conditional interest in acquiring the company, conditioned upon the successful demonstration of Scrypt and edge AI ASICs at commercial scale. Can you elaborate on what it would take for this interest to solidify into a formal offer or timeline?
Sure. I'll start. You want me to start, Yngve, or do you want to take lead?
No, that's fine. Yeah.
No, so to me, it's just natural that when you develop the world's leading technology within any vertical, large incumbents and other industry actors would take a look at you and have an acquisition in mind. That being said, I think Yngve might also have elaborated on this, but the board made a decision that this approach specifically was firm enough to be considered insider information, and therefore we needed to inform the market of it. How much, how large the stakes are, how much bravery they want to show in a potential hostile acquisition, that's up to them to decide. I won't really comment on it. I just don't want it to divert us from the end goal here, which is creating our chip.
Stay away from that type of noise until a formal bid is on the table, where I'm sure that the board of directors will consider that bid, and give the recommendation based on a formal offer.
Yeah. I can add that, when discussions like that start and we are approached and a conversation like that is started and, well, we have to read very closely what is the communication really about. And at this time, it fell under what we think is the regulations, what we are obliged to report. It's not really a choice. It's the rule book of Euronext. And when you're talking about, specifically when an acquisition is mentioned and it also the condition to arriving there is on proving the technology in a demonstration. It's sort of, we believe that our chip is severely de-risked when it comes to the technical and theoretical risk of the design. So for us then, well, we made a judgment and this is probably the best to be on the safe side to report to the market.
Yeah. I guess this question is probably impossible to answer, but I can read out anyways. If the possibility of selling all or a part of Lokotech arises, do you have a ballpark valuation in mind that you will consider acceptable?
I suppose that question is with regards to the subsidiary of Lokotech. I think the valuation of that entity would severely depend on who's the buyer, and what could the buyer bring to the table that we don't already have. But valuations aren't. I'm not going to even delve into that because we're not there yet.
Yep. Okay. Let's go from here. Guys, for you that are watching this live webcast, please use this chat to ask questions if you feel like we have missed something. Let me go to the. Have you selected a chip manufacturer and secured the necessary production capacity and delivery timelines?
Yes, we have selected a chip manufacturer. The production capacity and delivery timelines we have secured at least for initial production run and our estimate based on the presale, and they can be scaled up or down as presales evolve.
Could you disclose the name of the chosen manufacturer?
Global Foundries.
Okay. Yeah, I think we have gone through all of the relevant questions. Have I missed something for the viewers and for the management?
There was one on the mining industry in Norway. Did we cover that?
Please submit the question or in the chat. Please submit it.
Point eight, mining industry in Norway. I'll ask the question, Yngve. How do you view the long-term prospects of the mining industry in Norway, especially regarding political acceptance and support?
Yeah, I think that one is sort of a little bit. It's a good question and also interesting question because, you know, constantly you see in Norway, in the media, and there seem to be like something political about running a blockchain. And for me personally, because we haven't discussed it in the board, but obviously we think there is a future. And after all, I want to remind everybody that operating blockchains, running a blockchain, as part of a network, decentralized, is basically having a big spreadsheet where the algorithm is deciding who gets to write the next row in that spreadsheet. So you can be 100% sure nobody is tampering with it. It is basically objective numbers and reporting in that spreadsheet. And it's, I think it's a fantastic technology.
The bit where you're talking about crypto mining, and the revenue of the people that are basically contributing or operating the blockchain, that is just a natural course, but it's not like the main event of this industry. It's just that you have to be that way because the operator that is running the blockchain or helping or computing and writing to the spreadsheet, if you like, you have nowhere to send basically the invoice for the power we consumed, for the equipment we used. So there have to be some synthetic coin that basically is delivered to the person from the algorithm. That is the various different currencies.
The coin and the blockchain basically comes hand in hand, and it's fantastic technology. I think it will be. It is important for not only, you know, Norway, Europe, or America. I think, everybody will eventually be embedded in technologies like this. Yeah.
Yeah. To add on to that, I feel like the current, or the administration that is going to step up in the U.S., they have voiced themselves very positive towards this industry. I think we have four good years' worth of political stability on the major markets. I'd be surprised if Norwegians and the Norwegian politicians want to repeat the major failure of trying to prohibit color television back in the 1970s. Like, this is dangerous. We need to prohibit it. No, let it mature a bit. Let us deliver our products with silent fans. There's no noise issue. Power consumption will come down over time.
By the way, standby televisions in the U.S. spend more power than Bitcoin. So combine that with the fact that power is very local, as we now can see with the cables going to Germany, still very local. You can't transmit it and use it freely anywhere, right? And a lot of that power consumption couldn't have been spent on anything else anyways. Add on the fact that people are now using the heat in a completely different manner than what they did before. I mean, there's spas on Manhattan being powered solely by mining. You want to pull that out of the whole ecosystem because you don't think that it sounds to me is a very narrow-minded way to develop your political agenda.
Completely agree. I just think it's, it takes some time for non-technical people to wrap their head around how is this set up, what is it doing. But basically, if you are not going to allow for a decentralized spreadsheet to keeping track of different things, it can be anything from cod, what do you call it, quota, to basically where is the skin coming from, you know, to a purse or to your boots, all these kind of things you want to keep track of. It's perfect.
This technology for that is perfect basically for the sustainability goals of the U.N., and because it's objective, nobody knows it. Nobody can, sorry, nobody can tamper with it. So, I think it's a matter of time before politicians also they can wrap their head around it and, and then it will sort of be implemented in a larger scale. Yeah.
So if there's any politicians listening, what I would urge them to do is rather regulate it so that you make sure that the miners tax their revenue in Norway rather than export the hash rate to Cayman Islands or British Virgin Islands and pay all the taxes or no taxes at all. Regulate that situation a nd I think the political landscape in Norway would change pretty rapidly.
Especially for the smaller rural communities where the majority of the power is created. I mean, that brings a lot of economic benefit if they're able to do that.
Yeah. I understand why they don't want a very loud, industrial site just in their backyard. I understand that completely. But when they then banned it and removed that site, everyone was surprised that their power costs and, more importantly, the grid fee increased 25%.
All right. Thank you. We got a question from Maurice here. Are there any updates regarding activities in the property bought by Arctic Core in May this year?
Sure. I'll just give you a quick rundown on the timeline, and we were expecting it to be operational quicker, but when you buy relatively rundown real estate that hasn't been operational since what, 2014 or so, there's always some unexpected issues to overcome, but we have now power, we can very easily achieve redundant power.
We have redundant fiber, and we now also have a redundant backup facility, so my hopes is that all the T's are crossed, all the I's are dotted before Christmas this year, and we can launch Arctic Core's more traditional data center business, which means renting out servers for storing data for email servers for that matter, and also privately utilized servers. There's a large need, and especially now with the political situation, where certain competitors are Israel-based, or data is stored, data that should have been secured is stored outside of Schengen borders. That's not a very good situation to be in. So I also expect some onshoring of those types of activities, and I do see the stock of the first service just here in the office. I think it should be up and running before Christmas, but.
Yeah.
It remains to be seen if all the manpower is as eager as I am.
Yeah. I think also, you know, adding some conventional data center operation to Arctic Core is just going to be great for it. And, the redundant location in Halden, it's also close to ready. So it's going to be a nice service.
Thank you very much. Anything else you guys want to add, at last?
There was a question regarding what drives the growth of PowerPool's hash rate in the latest weeks. I have said that we have started some strategic growth initiatives, and some of that growth is because of those initiatives. Some of them is just organic. All the growth until August this year has been organic. So I just expect that to continue. When those strategic initiatives really start to pan out, you'll see spikes larger than what we've seen in the last 14 days.
Yeah. So there are coming new algorithms on in Q1. And if I am or we are right about that, we think that blockchains will be used more and more. Well, there should be a nice growth prospect for it for PowerPool there.
Well, in addition to that, with the competitors' units coming online and actually having home miners become profitable because they are not profitable with mining Bitcoin, that allows more decentralization. And, you know, they, the organic growth, I mean, speaks words when you have one of your friends that's running a miner and then they tell another friend, you know [audio distortion]
It's breaking up a bit. We got some connection problems here, but
Yeah.
Yeah. I think it's good that it happened. Can you hear us, Ola and Ben? We have some connection problems there. I think it's good that it happened at last.
We have one hour in, so yeah.
Yeah. Yngve, maybe you can answer this question. Are there any risks of reverse engineering of the miners?
No. I would say, no because we have taken special precautions, specifically on that. We have implemented ways to protect the circuitry within the silicon. But, of course, it's not like we want to test that. So we are very diligent when it comes to security and protecting our technology.
Yeah. Thank you.
I mean.
Yeah.
There was reverse engineering.
There was a question on the, you know, how have we protected that technology? Is there a risk of reverse engineering? Maybe you know.
Yeah. There's a lot of things that goes into it, including the physical protections in the chip for microslicing. There's also the PUF that implies that the communication with each chip is encrypted. You can't reverse engineer based off the signaling. The processor is also a custom suite. So, even if you understood what the chip does, you don't really know how to replicate the processor. And then there's the fact that if you take this design or anything close to it to GlobalFoundries, they will reject you. So you need to then not only reverse engineer the chip, but also redesign the whole thing to fit another production model. By that time, we're deep into generation number two.
The costs of doing it is prohibitive, in the sense that we are not an interesting target for nation-states that have a limited budget of reverse engineering military equipment, for instance. So we are very confident that this is not going to be reverse engineered.
It would actually be cheaper to acquire us given the cost of doing such an event.
Yeah. So we've got another question here. Is company doing mining with prototype Hashblade currently?
Yes and no. We are mining currently with the prototype, MPW ASICs. But those ASICs are very much scaled down. It's a 255th part of the commercial design. So we are mining, but it's primarily for making sure that the board design is good, that the IO works, that all the other protocols work, and are verified before we step into mass production. I think it's that. It's part of rigging the value chain to be able to support the transition from engineering runs into commercial units very quickly.
Just to speak on that, that same prototype unit that we fired up on September 15th has not had a down day since September 15th when we started mining.
There won't be any visible revenues in account.
No, no, no, no. It's small. But it is solo mining.
Also, you mentioned on scale. I mean, that was tile one, the 1 in 255 you said. There's also a size scaling .
Yeah.
S o it's a lot, you know, gross. The full chip is a larger scale difference, so there's a bunch of those running, but it's not like they are. I mean, that is for test purposes.
Yeah.
So just the first, I guess, when we get those engineering wafers, we will start to get some, you know, scaled-up prototypes that we can start to run. Yeah.
In addition to that, we're also, I mean, we've already released some screenshots of our mining firmware that's continually smoothing out, and also we've released the documentation for the API if people want to build on that on GitHub for the miner. Yeah, I mean, things are going fast and there's really, I don't want to jinx us, but it's really not too much other than just waiting for the ASICs, really.
Yeah. Question from Frederick Lee here. Has anyone broken the encryption and claimed one BTC?
No.
No.
Anyone smart enough to try will realize it's impossible. So it's a marketing gimmick. And what happened? We'll have to spend way more enterprise in order and computational power to break it if you're even able to do it within the timeframe we gave. So.
I think when we were doing this, there was an encryption key that wasn't comparable to ours, but it took them $680 million to break the key. But that was just one key. So I don't think with Bitcoin was higher than that, then maybe somebody might try, but I don't think we have a problem at all.
No. No. It was never designed to be a risky business .
What is the yearly target from self-mining?
Do you want me to speak on that?
Sure. My job's adding that certain potential financiers, they don't want us to sell this equipment and only self-mine. So it also depends on who's financing us in the end. But our own targets, regardless of who, we can speak of.
Yeah. I mean, on our current trajectory, you know, we're wanting to sell and we're wanting to self-mine. The self-mining with the quality control units that we take off of, you know, each batch, we're looking to build up quite quickly to, I would say respectively, we'd like to get up to about 8%-12% of the total hash rate is what we're wanting to do. And that, once we hit mass manufacturing, we'll be able to scale that quite quickly to get there. And plus that, again, how it helps us out with advanced RMA because if someone has a broken unit, they can send it back. We send them a new one and then we fix it and put it in our facility.
Okay.
Yeah.
Guys.
They're all.
Yeah.
Yep.
Okay, guys. Perfect. I think we are finished for today. To the viewers, thank you. See you next time. To the management, thank you for taking the time to answer questions from our members and your shareholders.
Yeah.
Thank you for hosting us yet again. You've been an incredible support in this manner. Thank you for that.
Great to hear. Perfect, guys. Thank you. See you next time.
Thank you. Bye-bye.
Bye-bye.
Thank you. Bye. Bye.