Crypto Megan Podcast | How HIVE Digital Technologies is Disrupting the Crypto Mining Industry

If you have 1 terahash of mining capacity online, how much money will you get paid in US dollars? If you paid $18 per terahash for that machine and the network is paying you $0.04 a terahash a day, you can actually go $18 divided by $0.04, and that gives you your ROI in days.

Wow. I mean, breaking down these numbers like that, it makes, it makes it pretty tangible for people to understand, you know, their ROI at the end of the day.

I like to think at Hive, through our really judicious and analytical approach to crypto mining, we've figured out the most strategic way to deploy capital.

Welcome to the Crypto Megan Podcast. I'm your host, Megan Nielsen. I'm a high-end crypto and NFT portfolio and Web3 strategy consultant. I'm currently on a world tour as a Web3 keynote speaker at all the biggest Web3, tech, and fintech events around the world as an industry leader To educate anyone from beginner to C-level executive, and of course, to advocate for women in Web3. My mission is to inspire new people to get involved in the crypto and NFT space with quality, entertaining information they can trust as pioneers in this cultural and financial revolution, a paradigm shift. My name is Megan Nielsen, aka Crypto Megan. Here in Barcelona, and today I am joined by the CEO of Hive Digital Technologies, Aydin Kilic, who's going to share with us just exactly how Bitcoin mining works. He's going to break down the intricacies of the process and the input output and energy. And Hive is absolutely disrupting this space. Hive Digital Technology specializes in building and operating cutting-edge data centers, and by combining expertise in data center operations with Bitcoin mining, it'll fuel the advancement of Web3, AI, and HPC. by actively participating in the decentralized digital economy. So this is a Bitcoin play, it's a mining play, and it's an AI play. It's very interesting. So I'm very excited today. So let's welcome our guests and get right to it. The listeners are from Draper TV, so a lot of them have traditional financial background but not a lot of background in crypto, although they do want exposure, for which Hive is a phenomenal play in my opinion, which is why we're here today. So like, can you break down what is the difference Yeah, yeah.

So, so that's a great question. So the, the best way to look at it is how do the manufacturers price ASICs? They— when you buy a car, right, say you buy a Ford whatever Bronco and it has 300 horsepower, they don't price it in dollars per horsepower, right?

Right.

They're like, the truck is $50 grand, that's what you get, you get 300 horsepower, and this is the fuel mileage, the fuel efficiency. In crypto mining, it's very different. They price it in dollar per terahash. So when you buy an S21 Pro and it's 15 joules a terahash, that's 15 joules a terahash is the efficiency, right? It's like your mileage.

Got it.

But if they say we're going to sell it to you for $18 a terahash, what that really means is that miner has 237 terahash.

Okay.

And so what you do is the retail price or the purchase price would be 237 times 18, which is about $4,200. So in the mining industry, the Chinese manufacturers— you got Bitmain, you got MicroBT, and you got Canaan— they don't say the new S21 XP is going to cost you $4,200. They say it's going to cost you $18 a terahash. There's a reason why. And it's called hash price. And so if you actually look at this network statistic hash price, what it is, is it's the dollars per terahash that the Bitcoin network will pay out.

Okay.

Per day. So what that means is, if you have 1 terahash of mining capacity online, how much money will you get paid in US dollars? And it's a function of 2 things. It's a function of Bitcoin price, and it's a function of difficulty. As difficulty goes up, hash price goes down. As Bitcoin price goes up, Hash price goes up. So today hash price is about 4 cents a terahash per day. So what do you do with that information? Well, you go, well, if it's 4 cents a terahash a day and this machine has 237 terahash, that means you're making about $9.50 a day today. Hash price can go up, hash price can go down.

Right, the difficulty level as well.

Exactly right. So the really interesting thing, if you think about it, is If you are paying— a really quick cheat code is if you paid $18 per terahash for that machine and the network is paying you 4 cents a terahash a day, you can actually go $18 divided by 4 cents and that gives you your ROI in days.

Wow.

Right? That's 450 days.

You are the math guy.

Yeah. And people don't realize that. Like, people don't realize that you can actually take look at machine price and divide it by hash price to get your ROI. Now, that's on a revenue basis. It doesn't cover the cost side of the equation.

Right.

And this is where we talk about efficiency. So efficiency is really a physics equation, and it's measured in joules per terahash, right? So joules is a unit of energy. So how much energy does this machine consume as an input, and how much terahash does this machine generate, right? So it's like your miles per gallon, right? Right? So, sorry, so yeah, miles per gallon, right? So how far do you go for every unit of energy you consume? The fuel you have, yeah. And so what happens is the better the efficiency— a better efficiency is a lower number because it means you need less energy for every hash. So for example, if you have a machine that has 15 joules a terahash and this particular machine is 237 terahash, Basically, an S21 Pro is 3,500 watts, which is like about 3 blow dryers. Like, it's a lot of energy, right?

Right.

So what does that 3,500 watts mean? Well, when you get your bill from the electrical company, it comes in kilowatt hours. It's expressed in kilowatt hours. So if you're paying— I'm gonna make up a number. I'll use a different number. Uh, say you're paying 3 cents for electricity, 3 cents per kilowatt hour. We go 3 cents times 3,500 watts is 3.5 kilowatts, right? Just divide by 1,000. So 3 cents times 3.5, and then you have 24 hours in a day, boom, times 24. That means that S21 Pro will cost you $2.50 a day to operate.

Wow. I mean, breaking down these numbers like that, it makes it pretty tangible for people to understand, you know, their ROI at the end of the day.

Yes, exactly.

Because it's usually so nuanced.

It is.

Like, they don't know how to break this down. It is.

And so, how I try to explain the industry is to break it down to fundamentals. It can all be modeled as a physics equation, and that's what we do internally at Hive, right? That's why we've emerged as one of the best allocators of capital. I like to think that, you know, when you look at Jim Simons and, you know, Renaissance and all that stuff, and they've got all these math PhDs that have, you know, ostensibly cracked the code of how to, you know, quant trade in the market, and they get outsized returns. I like to think at Hive, through our really judicious and analytical approach to crypto mining, we've figured out the most strategic way to deploy capital. Now, we're not the biggest in the game, but pound for pound, I think we have the best performance. It consistently shows, and this is a function of we've got a phenomenal team, right?

Right.

You know, my CTO has been mining since 2011, right?

So really deep crypto background, blockchain background.

Very super deep crypto background and blockchain, but we've also got a stellar executive team. We have the leanest G&A per Bitcoin mined in the industry. Our CFO, Darcy, he's the longest-serving CFO in crypto mining.

Wow.

Right? Yeah, he's been in the game since 2010.

So you've got all these stellar recruits. You've got the mathematician/physics—

Yeah.

You, the expert in that. Then you've got Frank, the visionary tech leader, Frank Holmes, who's the executive chairman.

Yeah, exactly.

And then now you've got also this deep crypto— I mean, it seems like you have the proper equation. To build, I mean, a powerhouse here, but you have, and you mentioned that there are others that are bigger in the game, but some would argue that Hive at its current valuation is pretty undervalued for what you can put out.

Yeah, I think it's fair to say that. We have research reports from Northland that came out, B. Riley, these are investment banks in the US. I think our stock is a little over $3 US on NASDAQ. We're NASDAQ listed, as well as TSXV in Frankfurt. The Northland and B. Riley Research reports gave us a $5.50 price target, which I still think is modest. Stifel gave us a $9 price target, and Canaccord gave us a $6 price target. So I think what the industry is valuing right now is not unit economics, but rather scale, right?

Right.

The industry is convinced that it's a scale game, but, you know, I think you need to be careful when you look at scale, because it can also be a— the bigger you are, the harder you fall.

Right.

In 2022, right, Core Scientific, the biggest of the big, what happened? They went bankrupt, right? They were too top-heavy, a lot of G&A. And so we have the leanest G&A in the sector. Uh, we also look for efficiencies everywhere. So we are pioneers in optimizing firmware, right? So when you get back to that notion of a physics equation, if you can start to adjust the efficiency variable, now you can start to optimize that equation. And that's what we do too. That's how we optimize our ROI. But to do that with foresight and awareness, you need to understand the principles of it, and you need to understand the underlying equations that govern crypto mining. So that's how we approach it, right? I mean, it would be a lot easier if I just said, yeah, bigger is better, we order 100,000 machines and we're going to be at a gigawatt by next year, right?

Right.

Which, you know, some guys are doing it, but I mean, what do I know, right? Like, there's companies now that are multi-billion-dollar valuations, and it seems that printing a lot of shares, which people that come from the capital markets world, non-crypto mining, they look at it and they say, wow, like, the dilution is sort of off the charts.

Right.

And so I wonder how that's all going to shake out in the end. It's one of those things, you know, what do they say, our parents tell us, well, just because all your friends are doing it doesn't mean you should, right? So, we have a lot of discipline at Hive, which is good. Frank calls us the Navy SEALs. So, I think that it's— look, we've got $150 million in Bitcoin on the balance sheet. We've mined profitably every single quarter over the last 3 years. And why I say last 3 years—

It's pretty phenomenal.

It's, you know, 3 years ago, we had the last bull market, right? It's— everyone was mining profit in the bull market, right? That's when everybody raised all that money. Yeah, it's mining through the bear markets, right? 2022 was brutal, right?

Making it through, it was, it was much more brutal than a lot of people anticipated. So if you made it through, uh, you're one of those that could stand the test of time, basically.

And, and the, the funny thing is the, the halving has been a relatively soft halving, and people might, uh, be shocked to hear me say that. Why I say that is this: when you dollar denominate crypto mining economics in dollars per kilowatt hour, which again is really the comparison to your electricity price, right?

Yeah.

An S21 Pro right now is doing a little over 10 cents a kilowatt hour. Okay, there are new nice machines that you can buy and run that have decent profit margins post-halving. Go back to the last halving in 2020, there was basically one model that everybody had, right? And that was the S9. And the S9 was doing 3.5, 4 cents a kilowatt hour of revenue, which was brutal. Wow. That is— It was brutal, right? That's why so many companies were almost going bankrupt, etc. Um, and so you, you need to have that awareness. Now, hash price is at all-time lows. But now, this is a more nuanced consideration, and so for your viewers, if they want to pause and rewind—

They probably will have to several times. Yeah.

But if you can understand this very key principle, I think you'll know much more about the intrinsic principles of crypto mining than most people out there. Hash price is commodifying. What I mean is, it is going down with time as hash rate commodifies. So the S21 is over 200 terahash, you know, per machine. The S9, right, which is like the Model T for crypto mining 5 years ago, right, had 13 terahash. So through advances in technology, these machines that use a relatively similar amount of power, um, well, an S9 was 1,400 watts and an S21 is, is a little over double that, but It's, you know, 20x the hash rate, right? So my point is, is the new generation ASICs are making so much more hash rate that the network difficulty keeps increasing. Like, we're over 600 exahash globally, and as that keeps increasing, there's a fixed dollar reward every day from the block reward, and hash price is really dollars per hash.

Right.

So as the network hash rate increases, The hash price will decrease, right? Because you have a— you basically have a fixed block reward. Transaction fees, you know, vary a little bit. So if you look back at the last halving event, the hash price was way higher. So it's more intuitive and it's more instructive to look at your dollar per kilowatt-hour revenue, right?

Okay.

And, and so that is, that is the one thing that hash price doesn't look at is, um, the underlying efficiency of these machines. So, you know, where do we go from here? I, I'm glad that we're— we've had a soft landing this halving. I'm glad that you can buy S21 Pros and still be making 10, 11 cents a kilowatt hour.

So what do you mean by soft landing versus other halvings?

Oh, I mean, imagine all the machines on the market right now were doing 4 cents a kilowatt hour. Everybody would be like screwed, right?

Yeah.

Like, because now you've got these huge public companies with all this overhead. Um, it's really hard to find cheap power that— because like, then as a network grows, like, think of it this way. Um, once upon a time, right, there were companies that had maybe 2 or 3 cent power. As that all gets snapped up through the US, right, I'm talking about industrial-grade power contracts. We can go to utilities and say, hey, I want a power contract for 20 megawatts. Okay, we'll give you 3.2 cents, right?

Right.

You know, it's in the Midwest or whatever, and that all gets snapped up. And then people are looking and looking, and it's like, okay, the 3-cent power is gone. Okay, now we're finding 4-cent power. That's gone. Now we're finding 5-cent power. Now we're being creative and looking at different sources of energy. Now, the caveat to all that is some people are going up the food chain in the energy markets and going behind the meter and all that stuff. Depending how you play that angle, sometimes you have to invest a lot of CapEx, and now you're contributing to the power production side. That's— that can totally be modeled, but it advances the equation a little bit because now you have to look at what is that extra CapEx yielding you on the lower power cost.

Right.

But if we just do an apples-to-apples and say, okay, for all the crypto miners that found industrial power load where they could go connect to medium voltage from some utility and stand up 20, 30, 40 megawatts. Nowadays, it's more like 80, 100 megawatts. Power price, the incremental power prices that have come on have been generally higher, right? And so the average power cost of the Bitcoin mining network, I believe, just as a virtue of scarcity, will increase with time because all the guys that have the really, really good 1 or 2 or 3-cent power, that's all spoken for.

Right.

You know, beachfront property, there's only so much of it that's going to get snapped up, right?

In certain areas, yeah. And you're going to have to look for another area.

Yeah, exactly. So the thing is, I believe, and I've modeled this, I believe that the global average of the Bitcoin mining network right now is about 4.5-cent power. And I believe that the global average efficiency of all— if you took all the ASICs and treated them as one giant supercomputer, It's about 40 joules a terahash. So if we were mining in an environment where all the ASICs that people were running, like the new ASICs that you were buying, were only paying 5 or 6 cents a kilowatt-hour, that's what I mean by a hard landing, right? Because there's almost no oxygen in the room for profit, right?

Right.

But you can, as long as you have a bunch of new machines, you can still show good mining margins in this bear market, right? And that's where things get tricky. Once the American capital markets started unleashing billions of dollars into crypto mining, it was almost like a cheat code. That's the dilution I was referring to earlier, right? It's just like, oh, well, we'll just go raise more money. It's fine, right? We'll just go raise another $400 million. We'll just upgrade our entire fleet of machines. Have we paid off our old machines?

No.

Who cares? We just got to show the market that we've got gross operating margins, right? So I could have bought tens of thousands of machines. They could be 40% paid off, 70% paid off. Doesn't matter. Nobody's keeping track, right?

They're keeping track of what's being output. Yeah.

Exactly. What is your current quarter doing? And it's all forward-looking, right? It's like the rearview mirror is not just crashing through—

It's just not healthy.

No, I don't think so, but I mean, when you look at investor investor sentiment or shareholder value. Like, I kind of joke that the rearview mirror is not just cracked in crypto mining, it's like ripped out the window.

Shattered.

Yeah, just shattered. And it's just like, how big can you grow? And so if you could say, well, we've got 60% crypto mining margins, or sorry, 60% gross mining margins, which is a financial term, all the miners put their quarterly financials, and it's just, if you look at the raw cost and profit from mining, Never mind the corporate stuff, just look at your revenue from your data center operations and all your direct costs, electricity, data center staff, data center lease, et cetera, or mortgage payment.

Mm-hmm.

What is that profit margin, right? So, you know, for us, I think it was 35% last quarter and some of our peers were showing 50%, but there's 2 things. Number one, those, if you just have all new machines, Again, you can be showing a much better profit margin, and if you're just raising capital to constantly upgrade your machines so you have nothing but the best machines, you're going to show a better—

Of course.

Uh, margin, but you've destroyed capital because like the whole point, going back to my original, uh, commentary, is that you've got to ROI on your machines in that first year to year and a half and then free cash flow until that machine is obsolete in your And obsolete just means it's no longer profitable.

Which is about 4 years, more or less.

Yes, 3 to 4 years, depending on how cheap your power is and depending on when you got in and out of the market. So if you think of that, you go, okay, well, maybe if I'm really good at crypto mining, maybe I get back 50%, 70% more than what I invested. Otherwise, you just bought Bitcoin in the first place, right?

Exactly. And that's the whole point of, of like, how does this produce more and what— yeah.

So, and so So, but with the whole rearview mirror is broken analogy, people can just not worry about that and just say, okay, we've got nothing but the best machines, we're high profit margin, and we're growing, right? Because the best way to grow, like if you have a 20-megawatt facility with old machines and you upgrade it with new machines, you'll have more overall hash rate because remember, the new machines use less energy, right?

Right.

So for example, there's another little cheat code. If you have 1 megawatt, right, and you want to say, well, how many— how much petahash can I get out of 1 megawatt? And if you've got a J Pro and it's 15 joules a terahash, you basically go 1 divided by 15, so the ratio is 6.6, and then I usually—

6.6.

Take off about 5% just for electrical overhead, so you times that by 0.95. So the answer is a little over 6. So you can get a little over— I should give you guys the right ratio— 60 petahash. The best way to think of it is a little over 60 petahash a megawatt. If you had an older machine that was like an S19 and it had 30 joules a terahash, you could go So 1,000 divided by 30, and that would be 33. And again, I take 5% off just for electrical overhead. So you get about 31 petahash per megawatt. So my point here is simply, if you had a megawatt of S19j Pros, you would get about 30 petahash. If you had a megawatt of S21 Pros, you'd have about 60 petahash.

Nice.

Literally double the petahash out of the same electrical So, one way to show growth is to just constantly upgrade your fleet, and, you know, the financial analysts like that stuff. Now, the other crazy thing— this is the craziest bomb I'm going to drop, is— I know I've dropped a few really nerdy bombs, but this is a crazy bomb.

Oh, it's great. We want to know the intricacies of this.

Yeah. It's, um, when you look at the multiple that our peers are trading at, They are Anthony Powers. He's a really good crypto industry analyst. He's a CPA based out of the UK. I highly recommend you check out his blog. Just Google Anthony Power crypto mining. And I think he publishes on Compass Mining. He's got his own thing going now too. Anyways, Anthony came up with this term, the hype ratio. And really it's just what is your enterprise value relative to the amount of Bitcoin you mine per month.

Right.

From the monthly reports we all put out.

Right.

Now, that's a powerful metric because never mind what your reported hash rate is, everybody's uptime is all over the map. You'd think like that maybe us, Bitfarms is a good performer, sometimes you got Iris in there too, but you look at uptime and the top 3 are maybe in like the 95 to 99% kind of range.

Right.

From there, it steeply falls off. Like, you got guys that are at 90%, 80%, 70%. Now, to be fair, sometimes companies are curtailing and selling power back to the grid and making more money curtailing, but that's not always the case. Sometimes they're just having uptime issues, or they're having a brownout, or transformers blew, or whatever, because they're not great operators because they've grown so quickly.

Right.

And it's just about scaling growth. And so when you look at the Bitcoin per exahash, that's another metric that Anthony measures, and by the way, everybody puts in their monthly production reports now. It tells you how good your uptime is. So how much Bitcoin are you producing for every exahash that you have online? So that tells you about uptime, and again, we rank in the top Consistently. But if you take that all off the table and say, I don't care how much exahash you're telling the market you have, how much Bitcoin are you producing per month, self-mined? And what is your enterprise value?

And those 2 metrics are the core of what they should be looking at.

Yeah, exactly, right? How much Bitcoin are you producing per month and what is your enterprise value? And because everything else can be smoke and mirrors. Okay, whatever. Like, okay, you're doing this, you're doing that, you curtailed, you didn't curtail, whatever. How much Bitcoin did you produce at the end of the month? And you look at the enterprise value, our peers are trading anywhere from 4 to 8 times where we are on that. So, so their enterprise value per Bitcoin mined is 4 to 8 times where we are. That's a huge premium. Trading at a 400% to 800% premium of where we're trading.

It's massive.

It's massive. If they were trading at like a 20% or 30% premium, that's one thing. And that's why I bring up the gross mining margin question. It's like, okay, if these miners have a slightly better gross mining margin of 40% or 50%, but they're trading at a 700% premium to Hive, it's just like, how is that good shareholder value? It's not, number one. Number two—

Yeah, I think that— what do you think it is that investors, they're just not really deeply educated in the mining sphere, so they don't really understand how to understand that premium. Or as you're breaking it down to me now, it seems crystal clear. In fact, actually, I was early on to invest in Hive. I noticed Hive— Off the grid. Yeah, off the grid. So I'm primarily—

I should have brought you a hat.

I know. Yeah, I'm a Hive fan. Yeah. I usually just invest in— I love your merch too, by the way, the Bitcoin. That's awesome. I mean, it's awesome. You guys are doing awesome things. I picked it up early on just because I wanted some exposure in stocks and also my clients, they like just making things easy. So like Hive is not only a Bitcoin play now, it's also an AI play, which we'll get into later and many other things. But for me, I noticed the value prop there. Why do these investors— why aren't they able to understand this breakdown, this ratio? Because they don't know the deep intricacies of—

So like the industry experts, like the Anthony Powers of the world and other research analysts at investment banks, They get it, right? But I think there's a couple factors. So first of all, the ETFs that index crypto mining funds are algorithmic, and they just go off of liquidity and market cap, right? So scale for the sake of scale, liquidity begets liquidity. And I've spoken to some of these ETFs, and they've admitted that they don't have a team of quants that are like, oh, who's got the best value proposition?

Nobody's doing You just want to show good numbers.

You know, and it's just like, oh, if we trade these crypto miners as a basket, right, we can just have a crypto mining ETF. And like, I don't know, like, they're all pretty much the same, aren't they? Is the mindset, right? We'll just diversify our risk by having ones that have like bigger must be better, right? Right. Is kind of the mindset. So I think that as the industry gets smarter, But we just have to tell the story. We got to get more exposure in the US, uh, because the US— like, we do 4 million shares a day on NASDAQ. We do, you know, a tenth of that on the TSXV. So it's about getting the story out in the US. Also, having a US-domiciled address for your head office allows you to be indexed on the Russell 3000. We're Canadian domiciled, so I think that's another— that's another factor.

Right.

Uh, and, and really just, you know, Getting the research out there, now that we're getting covered— so we're covered by Stifel, Canaccord, H.T. Wainwright covers us, B. Riley, and Northland. There's a couple more banks that are about to do research on us, so those are the 5 that have published. Getting that out there and being like, guys, like, these are super smart guys in the industry, they're giving us $5, $6, even $9 price targets, it's them saying it, not us. And hopefully the retail crowd and maybe even the institutions are just like, wow, okay, this is— it's one thing for you to say, hey, you know what, here's why I'm great.

Right. It's another thing for outside reporters.

Yeah, exactly.

And analysts to say, hey, like, look at the numbers here.

Totally.

Yeah.

So, um, so that's all good. But yeah, we're super excited about our 100-megawatt site in Paraguay, which we press released.

Now, Paraguay, there's a lot going on in South America. I know some people are actually starting to look into mining by volcanic energy. Cathy Wood was just down there. She invested in one of those companies. I think it's called Sodium.

Crazy.

Yeah.

Not heard of that.

Yeah. Well, yeah, I've, you know, is it geothermal? I'm not like too up on all the details.

Volcanic sounds cooler.

It does. Yeah, it does. It sounds— he told me volcanic energy. And so, you know, since I'm more of an investor on the investor side in crypto and straight, so mining, I actually learned a lot now talking to you. And traditionally, you know, we've looked at the hash rates. As an investment thesis. So it helps to know a lot more about how the mining cycles work, because the hash ribbons and when miners are capitulating stuff, we line that up with our investment thesis of where are we at in the cycle. So now, getting that deeper knowledge not only helps people watching who want exposure not only to crypto, but also to the mining side of it. Like you said, if it doesn't make sense in the numbers, you just buy the Bitcoin. But you guys are— the goal is to outperform.

The goal is 100% to outperform, and the goal is to have an enduring business that continues to deliver shareholder value. We have 2,500— actually, 2,600 Bitcoin on the balance sheet now. So you get exposure to spot Bitcoin at a very attractive multiple.

And a lot of the companies sell their Bitcoin off, right?

Yeah, I mean, we do both. So we have a hybrid strategy where we sell some Bitcoin too. So before public companies, public crypto miners existed, If you are what I call a true school crypto miner, what you're doing is you're mining, you sell some Bitcoin to cover your costs, and you hodl the rest, right? That way, you're basically having a growing treasury of Bitcoin, and that's what you do.

Which is a really attractive model, though, your hybrid model, that they get to have that exposure to crypto as well, not just—

Because the idea, and obviously, in terms of selling some of that Bitcoin to cover your costs, You're also selling some Bitcoin to repay your original investment. So the idea is that by the time you've gone through a cycle of mining with those ASICs, you'll have more Bitcoin than if you just bought Bitcoin.

Right.

And that's kind of the whole thesis there. The AI piece— Hive is a big Ethereum miner.

Yeah, and you were called Hive Blockchain Technologies before.

Yeah, yeah, exactly.

Now you've transitioned to Hive.

AI Digital, yeah. Having blockchain in your name isn't helpful when you're trying to open a bank account.

I can imagine.

Who would have thought? I don't know. You learn so much in this journey we call life. So we rebranded to Hive Digital, but it did reflect our diversification because really, as we were so deep in the innovation side of the business, what we were ultimately, whether it's hash power or selling teraflops of AI compute, And the bedrock to that is data centers, right?

Yeah.

And so, we had our ETH mining data centers were different than our Bitcoin mining data centers, which are different from our AI data centers. So, we realized we are a data center company and we run different types of compute, right? So, the AI strategy was really birthed from the evolution of our Ethereum mining pivot. September 22. Unfortunately, and I'm happy to go on record saying big mistake that they went to proof of stake. You know, hot take, hot take, and that rhymes. That's three hot take, big mistake, proof of stake. If it rhymes that much, it's got to be true. And and so you know, I say that jokingly but seriously because the beautiful thing about the proof of work consensus mechanism is it denominates the value of your cryptocurrency in energy. So it has a real cost. It has a real value. It's not just codified magic internet money that you can just airdrop a billion new tokens, right?

Right. If it's not bringing in outside liquidity, which has been a huge thing for me this cycle, like, let's put on the big boy pants for crypto and let's get that traditional finance into the crypto rails and have it be useful there. Because otherwise, it's like you said, smoke and mirrors and house of cards. A lot of times, if there's no demand for that, if there's nothing being built on it or nothing like pulling in outside liquidity, it's a lot of recycled liquidity. Yeah. So it makes sense what you're saying.

Yeah. And like, I think stablecoins have emerged in popularity because they're apparently backed by, you know, US dollars or whatever.

Bonds or whatever it may be.

Yeah. And so, so that's cool. So those are, those are backed by something. But if you basically If you basically look at the top 5 cryptos now, there's Bitcoin, Ethereum, and a few stablecoins. And so, but you do have these other proof-of-stake things like Solana, and they're like, oh, but you can do more transactions. But there's no value to the underlying currency. What makes Bitcoin so immutable is the fact that you have gigawatts of globally decentralized processing power.

Yeah.

power validating transactions on that node every 10 minutes. And the physics, the sheer physics of a 51% attack in the Bitcoin mining network are insurmountable. Right. So I think that is, that is, that is very key.

But it seems there are 2 different value props, like the proof of work is more like the energy is right, the currency, but the others are more like software.

Exactly.

And so they're both valuable in their own ways, but it depends what your investment thesis is and why you're using one versus the other. And the reason I say that is, for example, like one of the companies I've been working with the last 7 months, they're processing telecom settlements on-chain with their coins. So they're bringing in huge liquidity from the dinosaur industry, which is telecom, because traditionally you'd pay up to 3% on each side, cross-border settlements.

Oh, wow.

So they came in and they said, you'll settle instead of 10 days, you'll settle in 10 minutes, which is very important, because they need to settle instantaneously, right? Because they're selling minutes and data, and that happens. So they said, hey, we'll charge a 1% fee on each side, simple. And you get to settle on our coin with the— so the telecoms buy their coin, and they settle in stables. And so it's, to me, reflective of how this industry is starting to mature and what crypto is made for in the first place.

See, that sounds like a good actual application. That makes sense. I'm not ashamed to say this, but I lost money on a few ICOs in 2017.

I think everybody's got that on their resume when they've been in crypto.

I just like, you know, like, oh, we're going to tokenize this. We're going to tokenize— like, the dumbest one I saw was something called Trial Token, and they were like, we're going to— and it was a scam, by the way. But they were like, we're gonna tokenize law trials, so you will crowdfund your legal bill, and then if you win, if the lawyer wins, you'll get like a yield on your tokens. I'm like, what are you talking about? Like, what? Like, how does it—

But see, that's not solving an issue. That's really just trying to create some sort of value prop, right?

It doesn't make sense. Like, how— like, what you're gonna Crowdsource your legal bill? Like, what are you talking about? Just so people could speculate on whether you win or lose?

Again, the word speculate, right? There we are. Which has its point. It has its place here in the crypto market. But back to this, like Zebu, the companies that are doing this, now they're opening this system, they're decentralizing, bringing in nodes so that people can help with liquidity, security. There's no one central point of failure. This is what blockchain is for. And so those nodes now, they're sharing 0.6% of that 2% of their profit, which was $83 million in the last 15 months.

Wow.

Based on those 2%.

What's this project called?

It's called Zebu.

Zebu.

So we've been around the world. We just came from Singapore. I actually told them I was going to be on with you today, and they were like, this is phenomenal. We love those guys. So, and so now they're sharing that with the nodes, and the yield from that 0.6% is anywhere from like 20% to 40%.

Wow.

Right? And it settles in stable every 15 days, not their native token, because they don't use emissions from the native to settle the staking, because then it's like the house of of cards, right?

Yeah.

You're printing. So it's like this mechanism to me, this is what, this is what this industry is about. It's about what you're doing, pushing forward the innovation, having a healthy balance sheet, helping people understand what the value prop is, and taking away those smoke and mirrors.

Yeah.

Being like, let's actually solve a problem and make money and be profitable. I know you said that one of your things is like maximizing shareholder value through and getting profit margin through your, through your mining. But not only that, the AI play now, Like, this is why I love Hive, because you guys have that value prop because you care about the entire health of the ecosystem.

Yeah.

Right? So I'm super excited about what you're building. And I want to go back to then the health of your system. So now we're talking about, you just broke down exactly how you can kind of manage and gauge where the profitability will come from based on the types of miners that you have and the energy you have. So going forward now, you have other products. propositions because you've gotten heavy into AI data centers. So let's go over that a little bit about the—

Yeah, so that was birthed out of the Ethereum, uh, basically after the Ethereum merge, we pivoted. And so our AI strategy was birthed out of that. Reason being, we knew that the merge was coming, we just didn't know when. And so the GPUs that you could mine Ethereum with are— you could pretty much mine them with like a retail gaming card, which was great because that allowed everybody— it democratized Ethereum mining, right? One of our techs at our New Brunswick data center put himself through, uh, college mining Ethereum on his laptop, right? Ethereum actually had way—

That's awesome.

Sorry. Yeah. Oh, it's— I love it. And Ethereum actually had way more miners because you didn't— it wasn't as industrialized. Like, it was much more retail, like anybody could mine at home. And so the point of that is just to illustrate the native culture of Ethereum mining being super decentralized, but also those cards weren't data center-grade, high-performance computing GPUs, right?

Right.

So Nvidia also makes GPUs that are not meant for gaming laptops, they're meant Industrial grade. Stuff like Exxon. So ExxonMobil, they've— I think it's called Discovery 13. It's one of the most powerful supercomputers in the world. It's like a cluster of 2,000 A100s, right? And what are they doing? They're analyzing seismic data, right? Because they're drilling terabytes of seismic data. That's high-performance computing, right? So what you need when you're doing that sort of stuff is really high memory bandwidth, a lot of virtual RAM, And high flop speed on the processing side. My point is, is that all GPUs are not created equal. So what we did was with the Ethereum merge, it ended up happening in September 2022, but in July of 2021, we announced a landmark purchase with NVIDIA of $66 million of their GPUs. And these were data center-grade GPUs, right, that had capacity to more than just mine Ethereum. So we were like, okay, when we buy these cards after the merge, we'll be able to repurpose them into other revenue sources. And that's what we did, right? So first of all, immediately after the merge, we pivoted into altcoin mining because the cards were already set up for mining purpose. And so what we did is we'd mine these altcoins and we would not take custody of them, but we'd get paid in Bitcoin. And so it was—

So you converted Exactly right.

And so we went pure Bitcoin after the merge, and that for us was good because those coins, those altcoins, were generating under $1 per kilowatt-hour revenue. Again, that's how you do your apples-to-apples comparison, not just on the cost side, but how you compare different revenue streams, right? That— those altcoins were providing us $1 per kilowatt-hour revenues. They're about 70 to 80% of what ASICs were paying out that time. So it was pretty good.

Interesting.

By the way, Ethereum mining was paying out 4 times what Bitcoin was paying per kilowatt hour. So keep that in mind.

So yeah, now I see why you also preferred that.

Oh, totally. That's why I dropped.

Yeah.

And, and so, um, and so we were like, okay, this is good for now, but what's better? And so what's better was getting into HPC and AI. And that's also when the AI boom started taking off. So we started a beta test. We took like 500 GPUs, we listed it on a market marketplace, and we hit $1 million of annualized revenue after a quarter, and we're like, wow, this beta test worked. So we grew that to 5,000 GPUs, and then— so roughly 10x— and then we grew that to $10 million of annualized revenue, which we just reported in our June 30th financials. So $2.6 million for the quarter, annualized is over $10 million for the year. So that's, that's, um, how we wet our beak, and then we realized, okay, so we've got these NVIDIA GPUs. We want to be on the cutting edge. So we went down to the NVIDIA GTC, which is their big conference. We met Jensen. We had a longstanding relationship with NVIDIA, but really started digging in and we started doing AI test benching, loading large language models on our nodes. And we started installing those 5,000 GPUs. They weren't in our crypto mining data center. They were actually in a tier 3 data center. And so that was really awesome. By the way, I've got to get back for a panel.

He's got to get back for panel. We can't— yeah, we'll wrap up with this now, how you've gotten into AI.

Yeah, and so what happened is the tier 3 data centers we set up in Montreal and Stockholm, and that's where our GPUs are running doing AI. And so to do that, you also need to make sure the GPUs are running in servers that have stronger processing So we partner with Supermicro, with Dell, we've built Juniper Networks systems to tie it all together, and that's how you bring on compute for AI computation. So we're really excited about the future of NVIDIA's Blackwell platform. We see that that's really going to push the limits of what we're able to do. We've bought some H100s as well. We bought 96 H100s. 700s last December. And so we're— we want to grow our fleet of next-generation AI compute. We've got a lot. We really proved and learned so much converting our, our A-series cards, those 5,000 GPUs, right? And now it's time to really take it to the next level with what's in store. But we haven't said too, too much, so that's a good way to kind of leave them just—

Leave it on the hook for wanting more. And yeah, AI is most definitely the future, so people investing in Hive, then they get they get crypto exposure, mining, and now AI plays, right? So that's so attractive to me as we transition and everybody's sort of flocking toward AI. They get like the best of all the worlds with you guys.

Yeah.

So, yeah, I think that's— it's been really interesting, you breaking down all these metrics. I know I learned a lot on this show.

Totally.

Because I haven't been deep into the mining sphere before. And I think a lot of people will be very interested to see where Hive goes and potentially getting exposure to these things through Hive.

Yeah, so we'll— you know what, uh, maybe in like 3 months or something, um, just after Christmas, or just after Christmas, like maybe early January before Christmas, uh, it would be great. We could maybe do a follow-up because we're, we're cooking up some pretty cool things, and I'd love to say more, but I can't. So I've just, just kind of, uh, described about the direction that we're going in, but once we've, um, you know, gone down that path, um, It'll be fun. And I'm really looking forward to going down to Paraguay in the next few months as well.

Oh yeah, lots going on down there.

Lots going on down there. And I think it's a wonderful country and we're super excited about that. So yeah, it's a good time. And most importantly, on a closing note, I think historically after the halving event, usually anywhere from 9 to 12 months is when the bull market comes.

I was going to bring this up just as we close this. You guys just published something. I saw it on Bitcoin Magazine. That you know how the price typically appreciates for Bitcoin after elections.

That too, that too. So that's why I think it would be really good to do a follow-up maybe in December. Hopefully. Yeah. When we're booming. Yeah, yeah, yeah, yeah.

Yeah. Very cool. Totally. Hey, Aiden, thank you so much for being here today. Yeah, no, it's been great. Really amazing chat, and I can't wait to catch up with you three months from now potentially.

Sounds good. Yeah, let's do it. All right, guys. Thank you. Thank you.