Hi, guys, and welcome everyone to this special fireside chat with Lokotech. I'm Olav from InvestorWeb, and I'm excited to be here today to guide what we hope will be an easygoing and insightful conversation. Today, we are taking a little different approach. It's more kind of a T-shirt on the sofa kind of chat, where CEO Ola and Ruben will walk us through, from Lokotech. They will walk us through some updates on the Lokotech in a more relaxed setting. My role is simply to step in with the questions now and then to keep the flow going, but we are here to let the conversation evolve naturally.
So if you have any questions, you can submit the questions to me or send it to me through os@investorweb.no, or through the comment section on YouTube, Facebook, LinkedIn, or yeah. With that, I give the floor to Ruben and Ola.
Hey!
Hello. Thank you, Olav, for the introduction. I've been challenged both from clients and from shareholders, "Why can't you show up in something else than a suit?" And I figured this would be a good time to try it out, at least.
Yeah, it's a bit of a different image than we did last time, the last InvestorWeb we did when I was back in the office.
Sure, and a spot for you right here, Ruben. A little bit busy travel schedule.
Yeah
Straight from the U.S., reporting in from Madrid, huh?
Yeah, right here from the home office at this time, and trying to make the best of it.
Yeah.
Catching up on a bit of sleep from the jet lag. That's pretty much it.
We're all happy that we don't need to rent an office in Madrid for you alone. Congrats on that one. So we had a lot of emails, you know, coming our way, questions revolving around different topics, and we figured it'd be a good idea to have a small chat about it. There's no formal presentation, there's no graphs or anything. It's just the two of us. Primarily, the shareholders and other interested parties have questions about three topics. It revolves around the ASIC and the financing of the ASIC, and how confident we are in the timeline of that project. It also revolves around PowerPool and the growth aspects of that venture.
And the rest, we were trying to put in a kind of like, what about everything else that we're doing? Separate line item for our sake now. So gonna try and keep this conversation about those three topics. Start out with, I think it's probably the most important one, the Scrypt ASIC, the transition to PowerPool, and the growth prospects we're working with there, and then, you know, catch up the rest in the end.
As this is gonna be a casual conversation, if questions are appreciated, so we'll try and answer as many as well as possible.
Yeah, fire them away to, towards Olav. He'll shoot them, shoot them in when they're, you know, due. So let's get on with it. I think the first question we wanna really attend to is the technical and then the financing part later on. Obviously, it's connected to each other, but, anyways, we received a very positive noted question, and I wanna start with that one. And it reads like this: "In notice to the market, dated the fifth of September, Lokotech claims technical success in your mission to design the world's most energy-efficient Scrypt hash chip, ASIC.
Are you content with the process, and what are you particularly proud of having achieved?" So we'll just refer to this as the ASIC in this conversation, Ruben, so let's not use the long name, Scrypt ASIC chip.
Sure.
But if I'm content with the process. Yeah, sure. First of all, I want to address the last part. What am I proud to achieve? Well, first of all, we have beaten a 40 to 50 billion dollar company on efficiency. It's the world's most efficient ASIC for Scrypt hashing, and that is a huge achievement, especially seeing that it's on an older, more mature production node. And I understand that the shareholders have been, you know, tired of waiting and seeing delays, and what have you. But delivering a commercial chip design in 4 years at a very reasonable cost, and I know of competitors trying to create a cheaper, more easy ASIC, for SHA, for instance, spending $30 million , and they're getting nowhere.
And we're doing this in four years, which is a lot shorter than the industry norm of fabless semiconductor firms, which is, you know, between five and eight. And in this context, I'm actually pretty proud of the timeline. And my chairman and our external board members in the Nordic Semiconductor will vouch for that, and they're particularly happy seeing that we've come so far in so short amount of time with so little money spent. Which leads us to the next question, I suppose, which is, that there's been the timeline, yeah? There's been so many adjustments to the timeline, and like we're a bit confused. Can you reason around this? Can you explain? Sure.
Without digging into the details of, you know, personal tragedy or, the redesign and a little bit of misaligned engineering efforts on what ultimately became a redesign, I think it's important to remember that we are a small company, which means that everything that we do, every small delay, every adjustment, very quickly will be encompassed by the MAR law and market abuse regulations, right? So we have to report it. Usually, when you have a chip design effort, it's either a small private company where there's no news, no one knows about any delays. You don't realize they even exist until the fact is there. Or it's a large cap company that doesn't need to report everything, right?
There's a lot of delays and timelines shifting around, but no one really knows, except the guys working on the project.
Right.
So-
Even if you have, for example, a very simple chip, for example, if you have a SHA-256 chip that is already running on a higher node, like for example, the S9s for Bitmain. When they step it down to a lower node, that takes a very short time 'cause you already have proven tech, you already have everything. So that's essentially telling the foundry, "Well, take this design, we apply a few changes to make it compatible with the new node, and we just print it." It just takes, say, six months, 12 months tops. So there's not really any redesigns, any really big improvements that happen on the hash engines themselves, so it's a lot shorter timeframe. And most of these things happen behind closed doors.
You don't know that L9 was coming out until a few months before it came out. So it's hard to tell exactly when development started and when they, how they moved around. For example, the timeline between the S3 and the S7, they had the L3 and the L7, they had an L5 in between. We know nothing about the L5 except that it was not really commercially viable, so they just skipped it. So all these things, once you're a publicly traded company, you have to report them. So you have a lot of people that get expectations, and when SHA- things shift around, there's always, "What happened here?
What happened there?" So it's, there's a lot more inquiry and a lot more transparency, which is also one of the- I think it's the biggest strength of, publicly owned companies, which are small and are dedicated to the public, is that they- they're transparent even on the negative aspects.
Yeah, I mean, it's a two-edged sword, you know. Especially those very attentive clients that have followed us for a while, and like, "Where is the bulldozer? Where is the previous generation?" They don't follow us close enough to understand that it's the same basic, and it's, you know, communication targeted towards the capital market, not necessarily towards clients. So, it's a double-edged sword, but I think in this industry, very few are as transparent as we are, so it should work in our favor long term, Ruben.
Mm-hmm. Yep.
So the transition into the capital raise, I know a lot of people are excited about talking about that topic, and, you know, we reported so many times that there's NOK 60 million-NOK 80 million required to start mass production. It's multi charges. How will you finance it? The financing gap is getting closer and closer. Or you feel like there's something else we need to talk about with regards to being proud and happy?
Mm-hmm. Yeah, I mean, it's the financing gap is there, but there's many ways to approach it, and there's definitely, of course, I'm sure Ola will definitely know. But it's a lot of different avenues that we can go to be able to match it and start production.
Yeah, in my opinion, it's like there's five main avenues. The first one is the, you know, relatively simple one, one would think. You close a large enough prepaid pre-sales that you can bake the mask charges into the unit cost. You don't need to separate out that line item, and, you reduce your margins, but the volume is so large that it doesn't really matter for the producer if it's A or B. We're still working on that. Still in discussions with large, distributors and large miners, but, I think the planning for plan B or plan C has intensified a little bit, even though there's no, outspoken strategy, being decided upon in the board. That leaves us with the four remaining options, right?
And then there's the non-dilutive ones, which is the first one's a hybrid, where you structure a JV or a single-purpose vehicle financier, instead of buying the units, finance it with debt. The company places an order to Lokotech, and therefore, we can do the plan A. Bake the mask charges in the unit cost, and that JV will mine with the units with a, you know, shared ownership. Very often typical in venture capital, those type of money in, money back first, and it all revolves around the structural debt, debt letter. The second one is to do the same, but with the mother company. You have a typical convertible note that will postpone the dilution into the future when you have a clear answer on what it was actually worth.
Can be done in this case, not the main role that we're trying to make happen, but definitely on the table. And then there's the dilutive options, where you either have a private placement towards, like, an industrial investor, either a semiconductor investor or a, you know, blockchain, cryptocurrency mining investor, or even on the far end, an AI hardware investor. That's something I think will strengthen the company going into the future, but it all comes down to, you know, the terms of the private placement. And the fourth option, which I think is worth considering, is a rights emission, where existing shareholders are given the right to partake at a certain price, and that right can be sold.
You know, it's a fairer option perhaps than the private placement, and I think that's something that should be very much in the minds of the board if we go down an equity route, right? But then again, I think very often that rights issues are most successful if you have strong financial investors already. We haven't started pre-selling. I know a couple of the larger shareholders that would like to partake, but you know, those are founders, so doesn't really matter when it comes to the broader shareholding base, so to say.
Mm-hmm.
Um-
Right. But then there's also the option to, like, explore different areas, 'cause what we're seeing so far is that the Nordic region is not very prone to crypto-related businesses because it's more of like the U.S., U.A.E., and Germany that are more crypto-related, that they like venture capital in that aspect. So, a lot of questions were why we're not exploring those avenues of other countries, and Ola will definitely tell you more about it. But of course, we already are exploring them as possible investors for it. It's not only Oslo and Nordic-related areas that we're looking into.
Also, another question we're asking, we've been asking, that people have been asking a lot is about the U.S. listing. What's the plan for it? Is it something we're gonna do, something we're planning already on doing? What do you think, Ola?
Now, so, first of all, U.A.E., as the United Arab Emirates, very crypto-friendly center it has become, and if they think we're not talking to foreigners, then they're mistaken, first of all. We are definitely talking to foreigners. I think, as you said, there, if you're looking for, and if you're going down the route of a, you know, private placement towards someone, industrial, there's very few industrial players in Norway, so we have to go abroad. The second question is regarding the listing, right? And we have a power of attorney for listing, OTC. All the documentation is ready. Everything's translated to English. Sponsor is ready. OTC Markets Group is ready, and they're obviously keen to execute.
When we talk to financial advisors in the U.S., they're more like: "Well, why would you try and now list OTC?" It's all fun and games, but you do this small raise, and for them, it's pocket change. Like $8 million, well, that's less than what they throw at a, you know, seed round. So, their advice would be, you know, complete this, get a second match pre-sold. We don't really potentially need the cash, but if you do need more dry gunpowder, then just go straight to Nasdaq. This is a Nasdaq case. It's not a OTC case.
I'm like: "Well, okay, I'd like to listen to that advice." But if we find a cornerstone or someone willing to go into a mining JV venture, you know, fixed margin, they wanna enter Northern Europe, wanna utilize our data centers, start mining on it, and they still want liquidity, sure, OTC would probably suffice. The financial advisors are, they are telling us, "Don't do it." A couple of board members have some, you know, stories of Norwegian small-cap tech firms that failed miserably in listing OTC, and, my answer to them is: "Well, they're not in blockchain, not in AI, so it's a hot field right now." So, yeah, I think that summarizes it in a sense. Still sitting on the fence.
There's no way back if you do it, so I'm ready to do it. Just a matter of doing it on the right motivation, not only, you know-
Right
lower and higher liquidity. Doesn't bring us anywhere. So,
It's also interesting that you bring up the Scrypt aspect of it, especially like just mining yourselves or mining your data centers. And a lot of questions are about the Scrypt business case. Like, what does it look like? Like, is it positive? Like, if we just bought this for ourselves, would we just be making money printing the crypto ourselves and not having to rely on customers? And I think that the answer is yes. If we went the just self-mining route, we wouldn't have to rely on sales or anything. We still have to bridge that financing gap, but once that's done, it's essentially you're compounding everything you earn because you have a very effective miner, a very efficient.
So you have no competition, so you cannot be outpriced in terms of electricity costs. Because Norway is not the cheapest electricity price in the world, but if you go up north, you have very, very cheap electricity, you have cold, so it's perfect for mining. So self-mining is a very simple solution for avoiding the customer market altogether. But if you're starting from small, it makes more sense to get liquidity by selling to customers and then also be mining yourself to keep up your very nice continuous profit on your balance sheet while still having a nice image and nice customers, decentralizing the hash rate in the whole world.
I think in crude terms, that's more or less what the objective is in medium to long term, to get the customers to join. Well, first close the financing gap, then get customers to join, getting the Hash Blades in their hands, and then self-mine ourselves. So I think the business model is very, very simple in that aspect.
Yeah, it is relatively simple. A lot of moving parts, obviously, to make it work, but you have the advantage towards the customers, that's for sure. So I was trying to, you know, make this as kindergarten-friendly as possible. You know, I want to be able to draw this graph with a crayon type of approach to it, and obviously, if we do a minimum order, right, of wafers, mask charge, minimum order wafers, we turn those wafers into hash blades, we start mining, and we do that in the most dilutive way, you're still in the green, in terms of share price.
Yeah.
Because you keep on reinvesting that over three years, the EBITDA is more than what you've invested. It looks healthy. In the minimum worst-case scenario, sell zero, and we have sold already, so it can't be that bad, right? It can't be that bad. Then you know, transition that model into, okay, let's put some glasses on, be a bit more realistic. Toss away the crayons, start with an Excel sheet instead. And you add on a bit of sales, and I've been challenged on this SpareBank 1 Markets', you know, prediction that I came with. Was a bit cornered into giving up _a sales target number. I sort of told them NOK 5,200,000, so okay, let's model it out on that one, NOK 75,000.
Some of them are on our own balance sheet, some is sold, different prices, some assumptions go into, you know, wafer costs, at volume, and it looks pretty good. The range is like, you know, $30 million-$70 million EBITDA over three years. So it's a pretty case. And if you start adding on the case that let's try and replace all the hash rate, the theoretical case, but we know the hash rate is there, and no machines are hashing that amount right now. So what if we, you know, go into this when it takes a little marketing force, we don't upcharge the clients so much, but trying to, you know, protect our own profitability, what would that look like? I'm not going to even mention a number, but let's say this overshadows everything else, right?
So, I think the business case is relatively simple in terms of trying to calculate it. All the numbers are publicly available. We got better data because we got the pool data than what the available data is, but regardless, it's on a precision level more than anything. And, now I look to the future with optimism, I must say.
Right. Especially 'cause now that we have access to everything regarding to the pool and regarding all the data, we have historicals and everything, it's not only that you can do ballpark numbers, you can get down to the specific and do very exact predictions on where Filecoin will be, not in terms of price, 'cause that's a bit more chaotic, but in terms of difficulty, you can really go and do those models, see where you will be, how much ROI you'll get per blade, what's our effective cost, our production cost, everything. And you can. We've run these numbers, and that's why it looks so good. Especially 'cause now the models we've designed are, I don't know if. I mean, this could. We can explain this.
They're stochastic models. Since mining is a random process, you gotta model them in a stochastic fashion with a drift, which is the market drift, plus all the variances on the mining itself, so you can go and make very exact estimations and really understand what are our possibilities, so you can make ranges on, well, if we are more positive, less positive, more pessimistic, and all those values and all those ranges look very good, so that's why we say that it's a very good business case, and it's even in crude terms, when you subtract all the complexity from it, it also looks very good.
When you add it back in, it sometimes looks better 'cause you make assumptions when you're doing a very crayon on the wall kind of mathematics. You make assumptions that are mostly pessimistic just to cover yourself. But when you go down into the specific, you can remove those assumptions, go down into the technical and really see that those assumptions maybe were too pessimistic. So in most aspects, it looks even better than when you just do crayons on the wall kind of math.
Yeah, sure, sure. It's nice to have more additional on the team as well.
Yeah, happy to be here, to be honest. Very interesting space.
So there's some question about confidence, right? And that it revolves around, on a scale from one to ten, how confident are you that you will bring the ASICs to market within the next three to six months? And it's a complicated question, it's relatively easily written, but difficult to answer. And I think, you know, first of all, and what's your definition of bring to market, right? Is it, having a demo unit, here you can buy it? Or is it, finishing the whole, yield optimization curve and being in a mass production? Regardless of that, three months is too short of a time. We won't be done in three months.
Six months is a lot more optimism, and I'm very confident if you add on a couple of months in the end, that it'll be, in what, in my mind, at least, is bring to market. Meaning that here's a physical product, how many you want?
Exactly. It's the definition, it's a bit, yeah.
Six is to improve, eight is safe.
Of course. I mean, if we're talking demo units, production wafers, packaged and tested by us, six looks okay. If we have on hand for customer, eight looks perfectly okay.
Mm.
It depends on how your definition of time to market is. But it's regardless of which definition you take, it's still very close. It's very different when you're two years away to make an approximation, than when you already have a physical design you could put in today on a chip and say, "Hey, here it is. You can mine on it already." I mean, it wouldn't be three months 'cause, like, it takes three months for the foundries to send us the chips. So even if unoptimized, what we have right now, it would take three months for us to have it on hand.
But, if we're just talking, like, on having an already physical design that we can print, it's easier to make an approximation than if you're just finishing the design, or you're halfway through the design, or you're still on the planning phase. So right now, all the approximations that we make are a lot more confident compared to what we could do a few months ago.
Sure, sure. And one question we received was, I'm like, okay, this is a God's gift when you're trying to do a Q&A. So open, you can really answer anything, but I think we're gonna try and bake it into this topic, right? The question is, what's the pipeline going ahead? And, well, there's a lot in the pipeline, but let's try and keep it somewhat at least. We're not mentioning all the details, but let's try and take the broader picture and what's going to happen from now until tape out, and then eventually launch of a presale based on demo products rather than presale based on technical drawings, right?
Mm-hmm.
So.
Yep.
As you said, Ruben, we could tape out today if we wanted to, but it'd be an unoptimized design, which I think any shareholders that have been with us for a while, it's not in their interest to just cut the string, say tape out today, and let's go with it. It's a complicated process because what you have in the chip design, a commercial design, it it's an RTL code, right? It's a couple of thousand hours. It's very efficient, so it's only a couple of thousands, not tens of thousands, lines of code that is being translated into a graphical design, a physical design layout, if you want. And you can do that with an autorouter with a press of a button and go. But we don't wanna do that.
We wanna make sure that the graphical design is also as optimized as we can have it. So for instance, timing closure, which is, you know, after a router has placed this design, created a physical layout, it's done so algorithmically. But if you're adding lag to the process by missing a clock cycle because one signal is waiting for another signal, you're very quickly dropping in efficiency. So you want to make sure that the timing closure is as short as possible. There's no delay, and you wanna get the clock frequency up by tuning this right? So on different clock frequencies, you'll have different types of lag.
Mm-hmm.
So this process can sometimes it's just a matter of moving the physical lines a bit, but it could lead to a redo of the RTL code and then reroute of it. It's a feedback loop going between one team specializing on physical design and one team specializing on chip architecture. That process is ongoing, has been ongoing for a while. We have one physical design, but we wanna improve it to the extent that we deliver a higher hash rate and a lower power than what we have publicly disclosed. So-
A happy surprise
It's an important one and after that feedback loop is completed and we're done, there's diminishing returns. You can't do more. It'll take longer time to get the improvements that and what time it'll take to actually do them, and then you start the tape-out process. You take that design, it's a graphical design. It's just like that one from MPW, only a lot larger, right? It's two hundred and fifty-six times as large, at least, and you take it to the producer, and they do what's called Design Rule Checking, DRC, where they check that the physical design is producible according to their machinery, right, or production node, and there's varying degrees of errors or faults if you want. All chips have them.
Some of them are critical, push back to RTL code, need to redo it to go again. Or it's just a minor thing that we can waive. It's not your responsibility as a producer, it's our responsibility. We know it, we acknowledge it, we accept the risk. When we did this DRC in MPW, we had a record low amount of waivers. Like, neither the design team nor the aggregator had seen so low waiver count. So I'm pretty confident that this DRC, seeing that it's what, 95% of the chip has gone through it once before, it's a repetition of something we've done before, will go smoother than what we are accounting for in our time. Hopefully, we don't know, but yeah. So DRC completed, waivers are written, signed off on, then the design is taped out.
You know, it comes from the old, old days when you deliver a physical magnetic tape to the foundry. It's tape.
Mm-hmm.
Right? The graphic file was a tape. So, that's also partly why we wanna, you know, announce the tape-out process has started, and eventually the tape-out process has ended. It's taped out. We're gonna try and report everything in between, we'd have no chance. So that's something I'd follow out for if I were a shareholder.
Mm-hmm. And then the steps afterwards is we receive the blades, package them. Well, we receive the chips, we package them, and then, once, they're validated through through every test, they're sent out. That's pretty much it.
Yeah. Yeah, there's a lot of questions going around, you know, first batch size, how many chips, blah, blah, blah. It's also important to understand that the yield, meaning the amount of chips that you receive from one wafer, in the beginning, it's gonna be really small, and it's gonna increase because the tooling needs to be calibrated.
Mm-hmm.
So the first 100 wafers is what we call risk wafers. We don't know if it's gonna be one chip or a thousand, or fifteen hundred or two thousand chips. They're all free of liability. Produce as many as you can. So getting an accurate number on the first batch in that sense is challenging. But we know that the production node is mature, so die yield of 90% after some number of wafers. So yeah.
You cut off a bit there, Ola. But yeah, it's gonna be a ramp up on the yield, and it'll get up to the high eighties and nineties as production starts.
Then it'll drop off a bit?
Yeah, but it's all right. I mean, a few words only. Don't worry, Ola.
No worry.
Yeah, and then we also have a lot of questions regarding the latest power simulations per chip and the hash rate simulations per chip. And yeah, it's an interesting topic that a lot of people brought up, like, with our efficiency per chip, didn't drop much. The hash rate per chip did drop a bit. So, people are saying, "Well, does that mean that you will have to include more chips per blade with a higher wattage per blade?" And the answer is, perhaps, depending on those last timing simulations, if we are able to squeeze a bit more efficiency and hash rate on those simulations, a significant amount of hash rate and efficiency, then that might be completely mitigated, and we will be back to specs.
But the difference is, it's not much. If we have to, instead of have one side of the PCB populated with chips, we have to have two sides, that will be overkill 'cause that would bring us our hash rate to 80% higher than the target, 'cause the difference is not that much to require the full two sides. But it's still within the possibility on manufacturing, and the power consumption will not be. We still think it will be within the range of 10% that we gave on the estimations.
Yeah. So on that topic, Ruben, you know, this power simulation is done on a fixed clock rate, as you said. I'm not gonna, I'm not gonna announce which rate, but tuning that clock frequency in the stage we are in now can only necessarily bring higher performance. There's no way it can bring poorer performance. So the final power simulations is what we are so confident in. Another topic you didn't disclose, which we haven't disclosed, and we won't disclose either, is the die size, the size of the chip that stems from the wafer. And that die size has also been reduced since the latest estimations that we had back in summer of 2022 .
So on a cost per blade calculation, there's not that much difference, even though the power consumption might be in the upper range and the hash rate in the lower range. We're still fairly well within the earlier assumptions on gross margins and what have you. So just as-
And with that efficiency and cost, you still are, like, way ahead the competition anyways. So it's,
Mm-hmm.
It's a matter of perfectionism at this point.
Yeah, but then again, we spent four years so far, and we want to spend four years and a month more to get it all the way to where it should be.
Exactly.
Yeah, there's also a question regarding how large is the first batch? It depends on how you look at it. In terms of what we have sold from Lokotech AS, with minimum order quantity of one blade, that business is now closed. We're still negotiating with larger distributors, larger miners, regarding larger purchases because we want to at least attempt to trigger that plan A when it comes to closing the financing gap. In terms of a minimum order of wafers, we can only order the engineering wafers, but then we're kind of stuck in a time loop again. What makes sense for us, I think, would be something that, depending on the yield that we mentioned, ends up in about 1,500-3,000 units.
If the yield ramps quickly, you'll get more. If the yield doesn't ramp quickly, you'll get less, but that's what we're talking about in terms of, you know, first order of wafers.
Another really interesting topic that people bring up is, what is the size of the self-mining on the first batch?
If there's a minimum order, right, then what we don't sell, that would be what we could self-mine with. Some of them are for Q&A, RMA. If there's any issues with client's blades, we'll replace them, fresh out of the box, but minuscule compared to what we can put online. And if we're binning the ASICs, we can also put the ASICs that are, you know, halfway there, produce a hash blades that doesn't meet the specification, but doesn't matter because we're running them ourselves. We don't waste silicon purely because the end product isn't as promised.
Mm-hmm. And yeah, then after this, we start batch two, which also includes a few nice new products, which are more of the conventional mining standards instead of just Hashb lades. For example, the Single Tube and Double Tube, which are the L3 and L7 equivalents of size in terms of how they look. But they perform a lot better, of course, because it's a lot higher efficiency. So for example, an L3 equivalent, which we're calling the Single Tube Miner, I think it's called ST1 in our inner talks. It's gonna be an eight gigahash machine at about 800 watts or less.
The Double Tube is gonna be a 16 gigahash machine at 1600 watts, or less. Those two double additional products are gonna be part of batch two. We're already in talks with extrusion mold for aluminum for the body of them, themselves. They're just gonna have the same firmware and controller as the full stacks or the Rack Stacks that they're gonna be delivering in batch one. It will all be the same, like, kind of ecosystem, where you have the full stacks, which are meant for or the Rack Stacks, which are meant for the servers spaces.
Then you have the single and Double Tubes, which are meant for the, you could say, industrial specifications. So for example, the size and the immersion cooling and all those. And then you have the hash blades, which are just meant for home miners or gaming PCs or GPU retrofitting for old GPU rigs.
Yeah, you used to be an old GPU miner, didn't you, Ruben?
Yeah, I used to make the bodies for them, the rigs.
Cool
And sell them to people here in Spain. It was nice. It was like, it wasn't much volume, but I have friends here that wanted to get in, and they didn't know how to build their own computer. So I would help them, I would advise them. So it was interesting to see the refreshment of the GPU mining era towards Scrypt, which is something no one's really achieved or ever even tried to achieve. So it's really nice to be part of the pioneering team in it.
Yeah, that's why we, you know, opted in for a PCIe form factor in the get-go. But I think it's important for the listeners to understand that it's more or less the same product, just packed in a different manner, different price point, different hash rate, same efficiency, if you don't count the support components and fans and what have you. So, in the end, it's a lot all the same.
Mm-hmm.
Right now, we're testing different OEM manufacturers or casings, and you know, support components for the Rack Stack. What is it really? It's 24 blades put in a standardized rack-mountable server. It's a matter of finding the right supplier with the right capability and volume, and that will support, as you know, Linux, Windows or mining specialized Windows, sorry, Linux distros like HiveOS, and what have you. And
Yep
Volume is, I think, a pretty vital part, and it's encompassing that into a single ecosystem where we have one software that will support any number of hash blades, that will talk to the smart PSU, that can also talk directly to the pool and the fleet management software, so that you're tying the client into this one ecosystem that they thrive in, rather than try and make things work with like, some halfway, graphical user interface-
Yeah
and some little hack to make it work.
Exactly. Like, for example, like, HiveOS, like the way they detect, miners, on their system, is that they ping every possible IP in the local network. So you have to specify the local network. We're taking a different approach of having it work the other way. You can have on our fleet management system, where we're just closely integrated the other way around. You have the machines ping the fleet system, so that you have a better system. Instead of having to go downstream, you go upstream for the whole information flow. So same thing for pools. Pools don't detect miners, miners talk to pools, and then pools respond in accordance, most of the times. So it's just a better.
The architecture we're designing for this whole ecosystem, it's more streamlined for less hassle, so it will be simpler and easier to manage on the whole network, from the miners to the pool.
Yeah, I think that's especially important if you're recruiting new whole miners into the ecosystem.
Exactly.
Because, you know, the miners that you're recruiting now, they're used to macOS. They're used to having things solved for them, not writing terminal code in order to make things work, right? So-
Exactly
Important, important part, but yeah, and will you be able to let them navigate it via the pool software?
Like, you mean the control the miners and everything from the pool?
Mm-hmm.
I mean, it's not out of the question. NiceHash had a similar thing when they had the OS, the NiceHash OS for GPUs, and now they're doing a similar thing for ASICs. But it's their own OS, it's not. It's interesting how they approached it, but it's definitely challenging and hard to apply on the whole system. But if you are only, you could theoretically control just Lokotech, or protocol-specific miners from pools.
Mm-hmm.
And the objective was to make the API system for our whole firmware open for that specific thing in mind. You can have additional extensions to the Stratum protocol, which we will also be adding into those specifications at some point, to be able to have pool access to the miners as well, if it's necessary or if it's enabled. So you can add additional features on top of it, since we control the full firmware on the machines themselves. So you can have, for example, our specification can be public, so pools can talk and specify additional stuff to the miners to make it more optimized and to be more effective.
It's not something that's currently developed in the current market, but it's something that now that we have, we control A to Z on. We have not really control, we have knowledge from A to Z. For example, I know the pools side and the connection side, then we have the firmware side, which there's the whole Lokotech team works on, and we have the hardware side, which is our whole hardware designing team. So if we have all the A to Z knowledge, then we can develop solutions that are more optimized, instead of just understanding the A or the Z.
Mm-hmm.
So that's why the synergies here are very important.
On the topic of PowerPool, unless somebody has any questions and answer right now, I think we should transition into that.
Sure.
I'll give everybody a couple of seconds to respond if-
Yeah, you have one question. I got it from Fredrik Lyngås .
Right. Hi, Frederick.
I can read it aloud here. "I appreciate the diligence and the perfectionism in optimizing the chip. Could you give an estimation of when we can expect the start of tape-out? Are we talking about a delay of four to five weeks or will it-
So-
start this side of Christmas?
We had this incident with our chief architect, right? His o kay, so if you have, if you're gonna draw a skyscraper, right, you can have a thousand architects draw drawings, but you need someone to oversee the whole thing. That's his role. He's a brilliant guy. Very unfortunate, this accident. Giving a precise estimate on how much that had influenced him is difficult. That's why we went with five to seven weeks. A couple of those weeks, it was, you know, drifting. Then you add in vacations, Thanksgiving and Christmas, so for us, there's no immediate reason to rush it the week before Christmas, because there's no one to, you know, play the ball back to us again.
I'm hoping that we can do it a good time before Christmas. I think that very early New Year's is more. I'm a lot more confident if we can bring it one week into January, but we're still within that timeframe, right? Mid-December, early January. Does it really matter? Not in my opinion. Obviously, everything that's done is. I'm more happy with things that's done. I think rushing it into Christmas vacation is also not a good idea now that we can freely pick and choose our own tape-out date. The tape-out process, I believe, will start internally, but I think the first approach to Global will happen right after Christmas New Year's. That's my personal take on it.
And then there's also the fact that the DRC period is a huge, not huge, I'm overestimating now, but it's somewhat of an X, because it can take anywhere between one week till multiple months, right? I'm pretty confident it won't take multiple months. We've done it before, more or less the same design. It took less than a month. If the waivers for, or the rules haven't changed for some reason, something that we haven't caught, then I expect the DRC process to take a shorter amount of time than what we are planning for. So the overall, e xactly time to tape-out is difficult to state. The time when tape-out will start is easier to state. So yeah, in my mind, we're last two weeks of December, first week of January.
Great, thank you. No more questions from me.
All right.
It's a PowerPool. Now we are-
PowerPool, all right.
Yeah, we're blessed with our founder and CEO. So.
I appreciate it, I appreciate it. Well, we did receive a few questions on long-term plans and how everything's going, like, what's the competitive landscape like, and how is the product mix, profitability, potential threats, going into the future? I think it's better to break it down into 'cause that's a big question. Let's start off on, like, the long-term plans, and I will also add short-term plans onto it. Long-term plans is the PowerPool right now has only been software. So it's been just pooling software, added to some blockchain developments and optimization and everything. So there's R&D on that side, but it's mostly pool software. So the objective is to diversify a bit more and also bring in the hardware side.
So we're exploring on-site solutions for proxies and fleet management software that would be installed on-premise for hosting facilities and all that. So there's, like, a whole avenue that we're going on long term. I think we've done some market testing and some exploring around different hosting farms and self-mining farms that are looking into joining PowerPool, and we also pitched this as a possible element in the future, and they were super excited about it. They really like our philosophy of optimizing, and the objective is to optimize also on the sites that we operate in. So that's one of the long-term projects that we're looking into.
Short-term, it's, we're looking to get some certifications to be able to work with some bigger players in the industry that are only- that require some certifications for pools. Currently, there's almost none of the pools in the world that have them. Foundry is the only U.S. one that has them. So we're looking to be similar to that and be able to compete with them in the enterprise level. Especially because what we're seeing is that SHA-256 is very industrial-based. There's not many home miners. It's rare that you see a guy with 20 or 30 hosted S19s. You mostly have a very big enterprises that have many petahashes of hash rate, that require a lot of.
First, they're very doubtful of people, so it requires an intro to them, and then they're very cautionary, and they really want to avoid new players 'cause they stick to the old guys, even if they're losing money by doing so. We've talked to a lot of them that are on F2Pool and AntPool, and what we tell them is that if you just switch, you make 3% more. So they are starting to test them out with a few machines, 'cause they really move slowly when you're that big. So having those certifications really helps start the process a little bit faster and everything. We also have some new agents which are moving around in the U.S. and Asia, which are bringing us newer clients.
It's hard to put an estimate on how effective that will be, but if people have been tracking the hash rate for SHA-256, in the past month, we've increased 40% in size. It's 40%, I think. Yeah, 40% in size just from these agents alone. So, even if they deliver a percentage of what they promise, they will be very significant in the development as the future comes in. And yeah, that's pretty much the plays on short-term, long-term. Then, like competitive landscape, it's like, said, we're seeing that the big pools attract big players, and not because they're better, but because they're bigger. And that's always been the thing in crypto. Like what we saw.
It was a long time ago, so I may get this wrong, but I'm not gonna get specific. So there was a pool that got 51%, not because they were better than the competition, because they actually started offering worse conditions just to try and persuade people to go to other pools, just because they were bigger. So they were the big names, so they, a lot of people just flocked onto them because they were big. And that's what we're seeing in the SHA-256 industry. It's hard to get a centralization when a few players control a very big portion of hash rate, and all the players that have the miners are very cautious and try to stay with what other people do.
So it's kind of like a fight between what should be there and what's actually there in the Bitcoin mining. But yeah, it's these efforts are trying to, like, disrupt this kind of standard currently. Then the product mix we got the different algorithms. Although we, we're in talks with a few farms that would like to add a few algorithms in. Equihash is one of them, so that's one thing we're trying to keep an eye on, as well as ETC hash, and also the new miner that's been developed by
Ethereum Classic. Just for viewers that's not that into it, Ethereum Classic is ETC hash.
Yeah, ETC, yeah.
Yeah.
Ethereum Classic. Before ETH went to PoS, ETC didn't. So ETC is the only one that's so big enough to be mined with the. I don't think GPUs are mining it anymore, at least not at scale. Now, there are ASICs already for it. So it's the one algorithm that we wanna target 'cause I think it's very interesting, the move that they did on ETH against Ethereum Classic. So that's one of them. Then there's the new Blake3 miner by Bitmain, which could be very positive to be added as well, 'cause it's also an ASIC and it's.
I mean, we wanna be able to have all, every kind of user to connect to the system, not just the big guys. I mean, at the end of the day, if you have a farm with many miners, you wanna have them all in the same place and not have them scattered around different pools, so it helps you on your accounting and your taxes and everything.
We'll bring it back to the main, main topic. Product mix, first launch will be Blake or ETC?
I think the first one will be ETC, and then-
ETC
it's gonna be Blake.
Yeah.
Equihash,
Depending on that one.
it's kind of a floating player. Exactly.
Yeah.
Depends on how everything else develops.
Yeah.
But yeah, those are the-
Weird algorithm, Equihash, because it has the primarily larger miners, very few home miners, but they're not that large.
Exactly.
So-
It's not such a big volume market.
No. But, you know, if the one client you're talking about, if he jumps on and wants Equihash, you know, what's the development time for supporting that?
It's one month, one month and a half, tops. It's a fairly quick turnaround in terms of software, especially the way PowerPool's designed this. It was originally designed to have one, only one algorithm with Scrypt at the start. But then, I turned around and said, "Why am I doing it just for Scrypt?
Mm-hmm.
It makes sense, but it doesn't at the same time, so I did a whole redesign. So that's why all algorithms now are super easy to implement. So it's like the whole stack is very flexible in that aspect.
Right. And as you know, we've, I've insisted this, but we talked about this a little bit back and forth, that is additional features. Like, you know, diversifying PowerPool by having more features than competition. Very little development and investment to get something that, you know, diversifies you. The reporting part, I think, is important because I remember back when we were mining, I s cattering across all the different exchanges, all the different pools, get all the data. One pool went down, so we don't have the data anymore, and you're trying to recreate the data to have a accurate accounting, right? And I can only imagine the hell on earth if you're a private individual mining and you're going to at least trying to tax it correctly, to have all the data in one spot, right?
Yeah.
So that's-
Exactly
where one feature came, and I'm sure we'll find more features like that.
Yeah, the reporting feature from PowerPool is, I think it's super, super useful for big players especially, 'cause if you're trying to make as much money as possible when you're mining, you don't only mine Litecoin or Dogecoin or Bells, you mine a lot of other coins if you want to optimize it. And if you have maybe tens of coins that you're mining, even in just one algorithm, even excluding all the Bitcoin and everything, just talking Scrypt right now. If you have ten coins, you have to track the prices for when they were mined, all the differences of prices that happened over the day that you were mining, that you're switching off and on, where you're mining at that time.
Even all that accounting, it's nearly impossible for if you wanna tax it correctly. 'Cause you cannot take the average price, 'cause you were not mining for the whole day on the one coin. You were mining many coins on that day, so it makes it impossible. So having PowerPool have the automatic revenue reporting monthly, that it's auditable, that you can actually, like, use as your proof that it's not been modified because you have a cryptographic signature in there, and have it all in one unique single source of information, it helps a lot for when you have to do taxes.
Yeah.
Plus, if you only have to hold one coin in your balance sheet, 'cause you can have auto conversion. So you can just receive Bitcoin, keep it simple, and have, and we convert everything on our end, then you don't have to worry about all the tax implications of having mined all the other different coins. So it makes it a lot simpler.
And if you add in a bit of trading on top,
Yeah.
Make things even more complex.
Okay.
So for the next question, I'm gonna read it. I'm not gonna. I'm gonna do something a bit unorthodox. I'm gonna leave the room because I don't want Ruben to feel like he's censured by sitting together with me. Okay? So the question for Ruben, can you talk us through why you decided to partner up with Lokotech, and what are your personal thoughts for the future prospect of the company, and regards to the ASIC and pool mining operation and any other vertical business areas?
Okay.
Get the answer, and I'll mute and leave the room, and
All right. Well, this kind of put him in the spotlight, but sure. All right. All right, it's the joining forces with Lokotech. It was not an easy decision because the
No
There were a lot of things involved, but I think it's a good way to explain it. I guess would be perhaps go to the start. I first learned about Lokotech when they were in Vegas. I watched the conference, I think. I don't know if it was live or like a few minutes after they were live, 'cause it was posted on Reddit, and that was where I was doing most of the part of the pool marketing at that time. So I looked it up, and I read. I listened to a whole presentation, and they said they wanted to build a pool, and they were doing an amazing, new, efficient miner.
So I was immediately intrigued, especially 'cause it looked like GPUs, and I was originally, like I said, I used to build GPU rigs for people. So I thought, "Well, what if I turn this also business into making their systems into like what they right now have the full stack?" So I thought, "Well, maybe I could make them for them and be a reseller here in Europe." 'Cause the pool was based in Spain, so there was already all the paperwork were established to be able to be part of crypto and act in a responsible way and a taxable way here in Spain. So it was a simpler pivot or like additional feature that could be added in.
So I thought, "Oh, let's see what would happen." I never really talked to them. I didn't even reach out, just in the back of my mind. 'Cause I was doing so many things at the time, so I couldn't really spend a lot of time thinking about it. And then, out of the blue, I received an email from them that they wanted to talk about the product, and they were looking for someone, some pool or some guy that was related to pools that could help them develop their software that would connect the ASIC to the pool. So I said, "I'm in.
I'm game." I mean, it really happened over a few months that I was a little light on work, so I thought, "I got time, and plus, I'll get to learn the industry better." And they really walked me through everything, everything they were doing, and that's why I really, really understood the com- how complex building an ASIC is. I actually investigated a bit like before, just when I saw Bitmain, I thought, "Ah, I can do better than them." So I looked into it. I knew it was complex. I didn't have much time, so I said, "Eh, I'll push it back." But then Lokotech came in, and they really explained how everything works, and everything looked so complex that it even exceeded the expectations of complexes that I had.
But it really opened my eyes to what the work they were doing, and I was really excited with working for them. It was just for, like, helping them out on developing this pool connection system. And especially 'cause they said, "Well, let's test it out on your pool, and you can also get a benefit of people learning about it." So I was like, "I'm game. Let's do it." And then from there, I pitched to them, see if we could join forces also on the pool aspect, 'cause they, on the presentation, they said they were working on a Scrypt pool.
They said, "All right, let's see if we can join forces here." And after a bit talk back and forth, we decided to join forces completely, and they joined in as a joint venture with me on the pool aspect. So that's how everything came to be. And all is back, I see. And so far it's been great. Growth's been exponential, and really, Lokotech's opened a lot of doors for me, and not only in contacts, but also in learning how things work. 'Cause once, if you just know the pool aspect of it, you know a small percentage of it.
If you just know when you plug and play a GPU, you also know a small aspect of it. Now that I can see the A to Z, really infrastructure and stack of everything that goes into it, it's super, super interesting 'cause it really opens your eyes to everything. And it's collaborations are limitless once you can have access to the hardware as well as the software, as well as the cloud. So you can really tab a little bit on everything, try and improve everything, 'cause the trivial solution is usually not the best solution. You know, there's Occam's razor, it dictates that it should be, but once you get deep into the weeds, it never is.
It usually is so twisted and so weird that it even dumbfounds anyone that tries to, like, actually go for it, but it's really a very interesting space to be in and so many aspects that can be explored, so that's why I joined forces with them, 'cause I thought it would be mutually beneficial, and so far it has been, and it's been amazing so far. That's pretty much it.
Okay, tuning in there. You look finished.
That's all right. Did you hear any of it, or did you actually walk out?
No, no, no, I think.
That's awesome.
Okay.
Yeah, that's it.
Is it all right, Olav, or am I in big trouble now?
No, it's correct.
You say it again? Yeah.
No, I was asking Olav for his third-party opinion, if it was okay or if I'm in deep trouble.
It pretty much concludes the PowerPool, like parentheses on this pretty much, but we do have other questions on the other business verticals of Lokotech, because PowerPool is-
Sure, but thought it might be a good idea to see if Olav had any live questions coming regarding-
Yeah, not currently. I haven't received any questions. I encourage the viewers to send in questions either through the email you can see on the screen or in the comment section on Facebook or YouTube. If you're watching this presentation through Finansavisen, you can send me questions through os@investorweb.no.
All right, so quick repetition of that one. So we have a couple of questions regarding ASIC Exchange specifically. I wanted to cover those before we move into, you know, a short commercial update on all the other verticals that we are operating in. I trust that Ruben gave you a pretty good and clear view of PowerPool. So ASIC Exchange, the questions are usually regarding, okay, that's a small order. Will you order more? Or, how can a corporation add business in other verticals? And, you know, we ask 36 MW and blah, blah, blah. Different, different topics. First of all, ASIC Exchange is a medium-sized distributor, right, of hardware. That's their main business. They also have a data center business, 36 MW across the whole globe.
And they also broker data centers and data center deals. So quite a global footprint. I like them a lot because it's so short to the decision maker. Vlad and his partners are approachable and very, very well known in the industry for being able to deliver. I mean, the places in the world he's delivered miners, I mean, you wouldn't be able to pronounce them again, even. And it's so rural that you, you're not trucking them there, let's just say. It's a helicopter or it's a llama ride. Like, basically, a donkey even could work, right? So very interesting guys. So what other levels are we or can we cooperate on? Well, first of all, they are involved in the Filecoin business.
They provided free hardware for us testing the sealing infrastructure, so there's a revenue share agreement there, and we might add on more business there, especially if the mining of Filecoin turns around to the better. And there's obviously also some synergies with PowerPool. I'm not gonna talk specifically on his operations. It's like, as we say in Norway, asking a Sámi, "How many reindeer you got?" That's basically what you're asking a miner: "How many machines do you got, and which algorithm, right?" So what I do know is that he's spread across multiple algorithms. Some of them we don't support in PowerPool. Some of them is very natural that we start transitioning his hashers into PowerPool.
Some places are so remote that he doesn't even go there, and there's no remote capabilities, so we can't do anything without flying out to Patagonia in the middle of nowhere, right? To try and change some things. So some of it is off the table. But for us, I think it's most important that he gets comfortable with the VIP referral agency deal that PowerPool has, so that he can bake that business into his ASIC trading business. So that the volume of ASICs currently run rate 3,000 units of S 21 equivalents a month, to improve his client relationships as well. So I don't know if Ruben mentioned it, but there's a possibility to white label PowerPool, so to make, you know, make distributors able to offer that service that we offer in their own brand.
That's something that we need to discuss with him. And yeah, he placed a small order, but that was the smallest order that we would allow that type of key account management to do. So he's turning around selling this to his clients, and I'm sure that if his sales effort works out, it'll up the order. So remains a bit to be seen on the other verticals. Filecoin, in and of itself today, isn't that promising. Not our problem or not our reason or fault, it's just the economics behind Filecoin right now isn't very good.
But if he, for some reason, suddenly comes over a batch of hardware that he wants to use, he has got nowhere else to send them than to us, so we can increase that, the revenue share of it. So yeah, we're working on multiple fronts. Which transitions us into the other verticals, Filecoin. And I'm happy that you've partaken in that project more and more, Ruben. I think-
Yeah
Some of the headaches that the current team got. But in the very end of the development project, Filecoin decided to switch frameworks, right, from Lotus to Curio.
Yeah.
You're all set now. It's full speed.
Yeah, it's been an interesting transition, especially 'cause it's gonna be a better software. The framework is a lot more robust and better suited for our operations, especially 'cause we wanna do Sealing-as-a-Service. So it's a lot better in that aspect, 'cause it's a lot simpler to do Sealing-as-a-Service with that new framework. But they're still ongoing, so there's a lot of back and forth with the developers to try and fix a bug we have discovered and help them out. 'Cause at the end of the day, the project is about making our operations better, but it's also helping the ecosystem as a whole.
'Cause if we make the ecosystem better, price will follow, and the tech of Filecoin is very, very interesting. The cost per storage is a lot better than what Amazon or any other centralized company can offer. So it really is a very interesting piece of tech, especially for distributed storage solutions, 'cause there are not many in the world. And so it's a very interesting use case for the hard drives as well, 'cause the hard drives are essentially static machines. Well, they're rotating, but they're not really active in that aspect of like same as a CPU is, 'cause you can make it do stuff, so but data stored in a hard drive doesn't do anything. So it it's a really interesting approach to mining.
I think it's the tech's amazing. It's super smart behind it. But it is that way if we help the community enough, the price will catch up, so the economics will look better for us as well. We try and bring in a better system and as a whole. But yeah, Curio is running smooth. We've just integrated the latest software version, which will theoretically, as soon as we can get it up, which is gonna be next week, to test out the new performance. It should give us 2x the performance that we have right now. It's very, very good in terms of improvement. Yeah, everything's integrated, and it's running smoothly.
All monitoring is up again, 'cause once you switch from the one framework to another one, you have to rewrite the whole supporting components. So we had to rewrite the whole monitoring system. We had to rewrite the whole uptime monitoring, the whole, like, schedulers and pricing as well. So everything had to move around a bit, but now, everything's up and running.
Yeah. So taking on my, you know, strategic long-term hat, I'm very happy that Curio increases our efficiency, but the downside is that it increases others' efficiency as well.
Yeah.
That being said, the cluster is live. It's mining. As it stands, there's storage going on. It's in the Bleikvassli, the Hall of the Mountain King. We have plenty of nicknames for that place internally. Mines of Moria is one of them. Regardless, so combination of the technical development wrapping up, I'm expecting a report on the two different use cases, Sealing-as-a-Service, where we think we have an edge. We used to have an edge. Hopefully, Curio didn't remove that edge.
We still have the edge.
Okay. Thank you. And there's storage providing or mining Filecoin, as you, as you might like. I think it's a nice mix into the whole, Bleikvassli data center, location. Not very power dense, but right now, the economics are not unfavorable for, for Filecoin miners. Could, you know, change in a heartbeat, and we're very, investable, if that happens. We have the space, we have the, the sealing architecture, we have, the manpower to, to man it. So I think, if economics turns around and something we need to pay more attention to, and perhaps, seek some, external financing of, because, as you said, it's a, it's a very interesting technology when it comes to distributed storage.
So if there's any faith in that concept of distributed storage in the future, Filecoin is the way to go, I think.
Mm-hmm.
Which brings us over to other, you know, verticals. VIC, the messaging, or it's wrong to call it a messaging app, it's an encryption app. The bold critics are calling it an OpenSSL wrapper. Well, it's not. It's compatible with OpenSSL, but it hopefully will be compatible with other frameworks as PGP, Pretty Good Privacy and other encryption software. But it's about giving the consumer the power to encrypt themselves and not be dependent on any third party for various reasons, like the E.U. legislation now up for discussion, where every messaging application needs to have a backup sent to the E.U. It's a crazy future that we live in if that will be the case, but even so, some of the messaging applications are, you know, exploited, vulnerable.
So this now put the tool in the right hands, and we can decide to use it or not. Another topic, the first field testers, you know, frontline in Ukraine and journalists doing journalism in the U.S. have given their feedback for some requested features and some pricing feedback. So we are relatively close to have a version one point one launch, different pricing, and in that regards, we've also started to probe how we can put some paid marketing behind it. So lower price, higher adoption rate, some paid marketing that's status quo today.
Mm-hmm. Then we also have the data center vertical.
Yeah. Slow-moving field, real estate development compared to this tech world. But in Kautokeino, we have received a quote for the 2 MW installation. We are exploring a 5 MW option, so we're meeting Statkraft, Statnett, sorry, this week. Working closely with the grid company to explore that opportunity since the battery factory never became. So we could pull the trigger on 2 MW . Could be a bit too small in order to attract a foreign tenant or a foreign JV partner. 5 MW is within what they can consider enough to establish themselves there. But one important part of Kautokeino is that it's levied from electricity duty, so it's about 1.6 cents cheaper than any municipality where we don't have that levy.
And if you're talking renewable energy, Europe, Northern Norway or NO4 area is the cheapest, and if you can levy the duty as well, then we're pretty competitive in terms of power pricing. So I have some good hopes for the future. Unfortunately, the local grid code, it's taking some time, and it's a time-intensive process to get all the way to go, but as soon as it's investable, I think we should see some action in Kautokeino. Bleikvassli, we're about to power the first container. You buy that type of a property that's been, you know, run down and not taken care of for at least seven years up to nine years, then there you don't really know exactly what needs to be done when.
We're bootstrapping this and doing it as cheap as possible, right? We can't just pay ourselves out of any issues. The one cable that has been cut a little pickle because it had probably a sensor on it. In order to remove the equipment, they had to cut the cable. That needs to be replaced, but other than that, I'm pretty certain that within 14 days, we should have a fully furnished container ready to lease out, racks installed and everything. It's just a matter of attracting clients to a more traditional data center usage. It's operating a Filecoin cluster today and has a lot of capacity on that front as well. I think the site really needs a more traditional data center business model to really yield.
So that's Bleikvassli. In Øksnes, there's some discussions regarding installing a 420-kilovolt line. That's a massive improvement to the project if that happens. In the meantime, they're finishing the detailed regulation, and we are in continuous discussion with the grid company regarding the amount of capacity we can pull from there. So that's Øksnes. Halden is a bit disappointing. Extremely cheap brownfield area. Long term, I have no issues with it, with my background from real estate development, but it was a lot more pricey to get a enterprise scale installation there than what we first would have thought, seeing that the troubles are in the building.
So we can move pretty quickly on, you know, sub 2 MW, but, if you're doing, you know, 20 MW, 50 MW size, that will take a bit of, both time and money. And in that sense, Kautokeino is more attractive for the end client, so quickly moves to Kautokeino for at least mining purposes. But then again, Halden can probably be sold, can probably be just held until, they eventually go forward with the redevelopment of housing projects there. So a little bit of unknown when it comes to Halden, but I think the other projects are developing pretty well. Yep. And yeah.
That's pretty much it.
Pretty much it. At least the verticals that we've announced.
Yeah. Do we have any other questions that have come up for him?
No, no questions, live.
All right.
All right.
So-
Any questions, Ruben?
I think I'm pretty up-to-date on everything, so I think I know pretty much all the answers that I might have questions for.
Yeah.
But yeah, it's a thing we're gonna see a lot of speed up in the, in everything that's gonna happen in, on the chip, as in the coming months. Especially 'cause there's gonna be a lot of things to be done. But it's. It's, I think it's looking very bright in the, in this, coming half of the next year. And it's gonna be a very interesting time to be part of Lokotech. Not only from a work, from an employer's perspective, 'cause that's me, but also from, like, the financial aspects. That's gonna be very good.
Did you, you know, dig into what type of growth those agents would, you know, potentially bring in, in terms of numbers, or you don't want to dare to?
I think it's not really responsible to say the values they said they quoted to me.
Yeah.
'Cause if that's the truth, it'd be amazing. But I mean, I try to be more conservative on expectations. So I think we can even if what they've done so far, which is only been a few weeks' worth of work, they've increased the hash rate, like I said, 40% on the SHA-256 space. If they keep at this rate, we could see very, very, very good numbers very, very soon. And I mean, like we said, we're already very cash positive, like, PowerPool works fine. There's We're never in the red, always in the green. So it's even if they increase the hash rate, it'll only get better.
It's not a scaling issue. It's not "We fear that if too many people come in, we'll have a problem." It's quite the opposite. We're willing for more people to come in, 'cause we know it will be even better when they do.
Yeah.
There's no really any roadblocks or anything that could become negative from them joining the team. And if they're not as effective as they say they are, since all deals are commission-based, there's not really any downside to us. It's just there'll be less people coming in, not any extra money coming out of us. It's really positive deals. There's only upside for anything that comes out of it. And they've proven themselves as being effective, so I think it's really good for the future. I will not dare to say the numbers they quoted me, but it's really positive for the future.
Yeah. But would you say that it would be irresponsible to try and claim, you know, exahashes within half a year?
Yeah, it shouldn't be that. That's one of the aims is to reach that target within the near future. I think half a year is near enough to be considered like that. So yeah, I think half a year, it's doable for this, so for the growth we're seeing so far. So if they deliver on what they promise, that's gonna be no issue, hitting the one exahash range.
I really, I really hope that the Chinese agent that I introduced you to and that we have some history with and back and forth comes through on this one, because yeah, he's been a long-term Chinese distributor and you know, knows all the clients, and it's a very difficult market culturally and language-wise to enter. And I-
Of course, yeah.
I don't know how successful it would be to have a Chinese version of PowerPool, but something I wish to think about, actually.
I speak a lot of languages. Chinese is not one of them, so I'll need someone to help me out with that. But yeah.
Yeah, but yeah, software. All right.
Yep. Perfect.
I think that's it. We don't have anything else that we've received.
Yeah, perfect. I think this was a really great session. Usually we have question and answer, question and answer, but I think this was a great session because you got a really nice dialogue. So thank you guys for showing up today. And thank you to the viewers. We will see you next time. Do you have, do you guys have something else to add? Last-
Thank you for having us. I think this platform is really a good way to get people to access information straight from the source. I think this format's really been good, especially for a crypto space, since it's good to be able to answer questions freely and in a more controlled manner, and still people can break in, ask, and we can answer. It's no problem. Thank you for having us.
Perfect. And guys, you will be able to watch this presentation as a recording on Lokotech's IR page at lokotechgroup.com, I guess.
Lokotechgroup.com, yeah.
Yeah.
Okay.
So you'll be able to see the recording in there, too. And you are also encouraged to ask question through their IR page on the lokotechgroup.com, and hopefully Ola will answer them on an ongoing basis. Okay, guys.
Yeah, I try. Every Wednesday, I try to clear out the Thursday.
Yeah. Perfect. Okay, guys-
Thank you-
Thank you
for hosting us.
Bye.
See you next time.
Cheers!