Debunking Bitcoin Energy Myths | Alexander Neumueller CBECI

JohnPaul: . [00:00:00] Welcome back to Digital Gold season two, where we dive into Bitcoin’s energy debate with alexander. Neumueller cambridge University’s climate and digital assets expert architect of the Cambridge Bitcoin electricity consumption index, and a crypto native leader bridging mining and policy with deep roots in finance, academia, and media.

Alexander unpacks Bitcoin’s potential to decarbonize the grid, harness stranded energy, and redefine sustainability while guiding global institutions the World Bank. Today we’ll explore mining’s evolving role, debunking myths, and envision a greener future powered by Bitcoin. Alexander, welcome to Digital Gold.

I’m excited to have you.

Alexander: Thank you very much. Thank you for giving me the chance and inviting me to your podcast.

JohnPaul: So you spent a few years in the corporate banking and asset management section. What inspired you to leap into Bitcoin mining in blockchain research? Was there a specific pivotal moment you remember, or did it slowly creep up?

Alexander: Yes, definitely the latter one. So it did slowly creep up. So basically I started quite a long [00:01:00] time in banking, had multiple roles. I was, starting basically, very early on in my life in more the retail than corporate banking. Then went into basically treasury, , , prop trading was at the end, but I did in the bank.

I went then on, , at some point worked for an asset manager where I looked more into this investment sphere. Basically, me getting into, being interested in Bitcoin goes back a bit to my university days where I was always quite interested, specifically in a economics , and there is some relationship, right?

but at that point it was really going back to, I think 12. So, so when we talk about Bitcoin at that time, it was very different that not is today, but how people perceive Bitcoin in a completely different world. And I need to admit, I was ideologically interested in this new technology, I was tech savvy, right? But not to a very large extent, possibly not to the extent that I would’ve been, because currently nowadays much information about Bitcoin [00:02:00] available, but at that time it was very hard to get anything. And then it was more kind and these things. And, , yeah, sometimes it was quite technically and, , yeah, it was also different environment in terms of specifically with, , yeah, how Bitcoin was being used at that time.

So I could ideologically somehow identify this is why I was interested in the subject. And, but at that time I was unfortunately not investing in it. But I kind kept tabs on it, so I was just, okay, let’s, I may not invest, but I, it’s just interesting to see how it evolves and yeah, I, I kept interested and at some point I started mining myself. And, bit, a bit after I did it, I started it, so not but doing it myself physically, but I was, basically, , outsourcing it essentially. And, , in thousand 19 I formed my own mining company, but everything was done fully, remotely and everything was set up in China. So, yeah. I basically, this was how I got more generally involved in [00:03:00] this space.

And at some point I went to Cambridge to study before my role at the center for Alternative Finance. And yeah, this time 2021, right? Was a bit of a tough operations in China. And yeah, it was do

finance. Do I look into the blockchain space, did some things there as well, specifically related to defi, , or do I do something different? And specifically looking into electricity consumption. So actually develop , our electricity consumption index. At some point I was taking it over and , made some adjustments, put on new content, right. But yeah, this is how I end up in the center because I think I had this, I would say expertise in how the industry works, specifically specific, which machines are used and these things, right? How the market works. And, I [00:04:00] tried to bring this knowledge to my work at the research center and work on improving specifically theoretical estimates that were out there at that point, which, yeah, how many did in multiple updates of our methodology.

But then, and I think one of our key points. I had, , seen recently after I joined, I, , had contact with the industry and over time this connections developed. This is how I got in touch with kind the mining council. But it took quite some time to, to find a bit of a common ground, right?

So how can things be, how can we leverage synergies between what they’re doing and between what we are doing? Right? Which, , early last year, fortunately culminated in the creation of this survey where we found that, well, , how deep can we ask U minus, right? What can we ask you about and making, but in the way that is in line with, our own kind internal guidelines

and.[00:05:00]

People who are invited to participate in the survey are really, participants in the industry, right? So it’s not just someone who, fills out some link that we got, somewhere from, and we just, everything that people provide to us know , every minus has been individually onboarded and they’re, and yeah.

And finally, , last week, all this work culminated in the publication of this reward. And yeah, very, very excited to be here today to talk about it.

JohnPaul: / So you mentioned mining in China and that’s some of the expertise you brought to Cambridge and updating the model. Can you talk about specifically maybe what you saw in your mining operation and then maybe what they were doing? Maybe not wrong, but. What you changed in, in this new, presentation or new report?

Alexander: , yeah, this was, basically, I did not directly invest, I started mining because , having this finance background you , unite cash flows and these things unite DCF models, and this is how I came into it. to me was a bit more [00:06:00] related to my past mining than just, , buying ling Bitcoin.

this was not, was interesting, right. But I think the mining space was a bit more related to my finance past and was more exciting and more interesting, frankly. And yeah, I think there were at that time definitely good opportunities, right? . We still had the entire supply chain essentially was in China, which was very convenient, because you had the manufacturers there.

You had , , the farms there, you had the repair centers there. So everything was pretty much, , established. And , if at that point globally, , the US had some hash rate, right? But this, in 19, it was still a completely different environment. , Everything was very, very much, focused on China.

And yeah, , the first thing that, , disrupted , my operations a bit frankly. , , And I’m sure where the s. , Kinda issue where with the heatings, , and specifically if you’re not managing yourself, you’re completely reliant on, , you, I had obviously trusted parties, but [00:07:00] still, with the repair sentence, if you can’t do it yourself, right, you’re a hundred percent reliant , and, , a lot of bad batches.

And this was the first thing. And, and in the end, , after 2021, specifically, if you are , if you were a , small, sized mining operation, , there was, COVID logistics was terrible, and it was banned. I was even happy that received a couple of machines, , that I could sell at some point in somewhere.

and yeah, I think it was, a very rich experience, if we want to call it that. And definitely lot of things I’ve learned and yeah, , currently how I transport this knowledge specifically to my current role, it’s just kind. a bit of the established network to a lot of folks, right, globally, which obviously is incredibly important for doing things this survey because you need to get miners to contribute, to provide data, but specifically private miners, right?

To have under no obligation to provide anything to anyone except your balance sheet or your annual [00:08:00] reports. and I think that these things help, it also helped understand talking to folks, , because I didn’t never know how it works actually on the ground. So I need to admit, I have never installed an asic. at some point, actually, I’m now, I’m now aiming to buy, I don’t wanna name any brands, but to buy some small at home mine to just try it out a bit.

to just see how it works practically. Because I always just saw, on your pool, your hash rate and things if something didn’t work, , it was , the operator telling you, okay, there’s this problem. And then you managing the logistics to get it to the repair, more having this high level view.

but still, , you still know how things work. And I believe that definitely helped me to bring this knowledge to my research.

JohnPaul: And so it sounds you have people that have mined before yourself, but maybe were disconnected from the miner. You have people that, have never mined at all, and then you have people that are minding and physically working on the device itself. How do those three different perspectives kind [00:09:00] of, how have you seen them change?

Not change, but how have you seen them perceive Bitcoin mining differently? , You’ve dealt with a lot of the headaches of a mining owner. Maybe not as a mining owner who can physically access his hardware, and then I said, versus someone who just has heard about Bitcoin mining but has never even attempted it.

When you’re in your research, did you come across, and as you talk about your research, how do those different people engage and view Bitcoin mining?

Alexander: Yeah, so good point. I think we, specifically in my work, you work with a lot of different stakeholders. And indeed there are some, , obviously miners who, , fully. , I would not necessarily say there’s a lot of difference in perspective between those who, for instance, only do hosting with those , who basically do everything , in terms of how they perceive mining.

I think they perceive it pretty similar. It’s more, I think, the, knowing the nitty gritty of all this on the ground operations that some know and some don’t, but generally people have a bit of an understanding, right? Even if you have never [00:10:00] built something on ground, that of IFUs air quote miners, , you need to make sure the air flow.

So it’s more , , the difference in my work comes down to speaking sometimes with folks who might have, , had no exposure to mining at all. Right? Possibly to having heard about Bitcoin, if even, , who definitely, but, , having also a bit of a, a background of what is actually Bitcoin. , And then even if that is there, , you might have folks who might have not really, even a broad understanding , of what mining actually is currently is changing.

And so over, over the time , I’m doing this work here in the research center, I think what we saw is definitely an increase in the understanding of the overall global community from all stakeholders. And it really goes from, , regular people. if I would ask someone on the street.

Do what Bitcoin mining is? most likely they’re gonna tell you add something with data centers. And something, , computing. And so it’s not the complete back box. You have a very, [00:11:00] very broad idea and they think the same is true for a lot of, let’s say, , regulators, policy makers, and these folks as well.

Where you see obviously more as the industry becomes bigger, in some jurisdiction, some , stakeholders and such, stakeholders are getting a much better idea of what it actually is because they might either want to promote it in the jurisdiction or do the opposite, right? Some say, well, this is not a.

So , you see this different perspective, but overall, I think, and , one of the points in the report is as well that it should serve as a bit of a foundation. The first edition where it starts really about , what is Bitcoin firstly, and then also explains mining, not necessarily in extreme amount of detail, but at least at a level of detail where those who go through it have some , I would say, basic understanding of, okay, what are we actually talking about here?

And yeah, this was also one of the, this educational component was, at least for the first edition, a big, thing for us [00:12:00] to include to make sure that, , someone who might be a newcomer has, if you start at page one and you finish at the last page, you know on the surface of , what is actually going on from what is Bitcoin, what is mining, what is the currency of the industry?

And then a bit about, okay, where things are heading.

JohnPaul: And so it sounds you have people that have mined before yourself, but maybe were disconnected from the miner. You have people that, have never mined at all, and then you have people that are minding and physically working on the device itself. How do those three different perspectives , how have you seen them change?

Not change, but how have you seen them perceive Bitcoin mining differently? , You’ve dealt with a lot of the headaches of a mining owner. Maybe not as a mining owner who can physically access his hardware, and then I said, versus someone who just has heard about Bitcoin mining but has never even attempted it.

When you’re in your research, did you come across, and as you talk about your research, [00:13:00] how do those different people engage and view Bitcoin mining?

And I did read part of the report, obviously it’s very long, but I did it how you set that up and how you, clearly state, here’s a building box. Before we jump into , mining’s impact, the head of, climate research for this project, for Cambridge Digital Asset Program, what’s the most surprising finding your team or yourself uncovered about Bitcoin’s environmental footprint or just about Bitcoin mining’s interaction with the rest of the world.

Alexander: I think . One of the most interesting finding for myself, right? It always depends a bit what you expect specifically if you’re very much engaged in the industry. There were a lot of things that might be very interesting for those who might not have daily engagement, or are continuously up to date with where things happening.

Personally, I think the, power coming from otherwise flat gas, the extent I think one of the, for me personally, one of the key research findings, because I think this is one of the best use [00:14:00] cases where, , really someone is coming in and off takes energy that’s otherwise really being, because if you just flare natural gas, there’s absolutely no economical point to do that and no point for the environment.

And if you can find, , this specific report was bitcoin mining, but there is also. HPC operations that you can build, but specifically given the CapEx, given the flexibility, I think, digital mining that regard is a very, fitting offtake, of this in, in this specific use case of, for instance, otherwise, otherwise gas and the extent of it.

And then trying to convert the extent. So it was about 500 megawatts. Convert this to, , extra emission reduction under different scenarios. Right. I don’t wanna go into too much detail here, but it’s in the report, right, where you have different scenarios coming global if you use different global warming, , potentials and, periods,, if you use different assumptions for combustion [00:15:00] efficiencies, and then in the end trying to, , show different kind scenarios and show that might, for instance be even a positive effect something.

on the. Even possibly a bit of a mitigating effect by just, , using a, a gas engine instead of a flare stack and possibly having a higher conversion of, , essentially, , methane to, to c2. And this is what I think was one of the most interesting findings, , that from the importance, I think what we is, the electricity mix was changing.

I necessarily those in space. We might talk about this a bit later, the, we captured a lot of hash rate, right? So we captured the 50% of the network, but still there is a bit of a US overhang, let say this. So in terms of jurisdictions close to 50%, where US located, if you look at the distribution of hash rate, also very US focused.

So definitely, but still, it,, shows [00:16:00] a direction because overall I think there’s not much question about the US being the predominant mining hub in the world. It’s more about the present. We go from estimates around 40% to our estimate, 75%, which personally, I think the one is maybe a bit too low, but the other one too high.

So it’s a decentralized, very hard. To essentially establish where folks are located. And you can try different things, work in pathways, mining pools now with directly, serving miners. And I think those are the ways with both the limitations and obviously, benefits. So, but yeah, so much, so much to that. , and sorry if I, if I digressed a bit, please.

if you want to drill into , any points deeper, but yeah, possibly another point was ewas the repurpose, recycling , and reselling that a lot of people I guess find interesting as well to see this. We have climate mitigation strategies where we show different [00:17:00] strategies, , to what extent miners are using those.

We have a lot of sentiment questions in there as well. So yeah, I think those are the, the interesting, the most interesting aspect and possibly to some extent also quite exciting.

JohnPaul: My next question is, how do you ensure that the Cambridge Bitcoin electricity consumption deck stays relevant as mining evolves? And did you guys identify any critical data gaps that still need addressing or things you’d to explore in the next version of this presentation?

Alexander: Yes, I think it’s a very, very good point. So first of all, we have now I think, two different, if I may call it product. The first work is more on our website and essentially the estimates you can see there are based on theoretical modeling. So that’s a bit of the difference to the report where if you go on the website and you see an estimate, for instance, for electricity consumption, it is essentially a daily estimate.

Everyday estimates are updated and we keep this. The findings we had [00:18:00] in the report and based on minor participation in the survey corroborate that , that our theoretical modeling in terms of electricity consumption works pretty well because even though we had two different methodologies at this specific point in time, which was the thirties June, 2024, we saw a very, both estimates in terms of electricity consumption from, , minor surveys and our theoretical modeling very closely aligned.

So , we see if we use past mining council estimates as well. So we see , there is an alignment even to different methodologies. So we are very confident in the electricity consumption estimate. This has been updated in 2023. To specifically account for periods of very high profitability because there wasn’t the past bit of an issue where our, I would say , first model definitely, , had some issues when it comes to periods of high profitability, but I think the new model works works really well and we keep it up to date by essentially having what we call a hybrid top down approach, which [00:19:00] means that, we use essentially on chain data to determine profitability of mining.

From that we form a basket of really exist of real world hardware or really existing hardware. This is based on the list that we are constantly updating. So if the new hardware models, we’re gonna take them into the list and this is , how we keep things updated. And yeah, I think in terms of electricity consumption, this works really well.

The difficulty is really how you translate essentially electricity consumption to emissions. If we keep things at the environment, if we look at the environmental component right. and there in the past you are generally reliant on so-called location based estimates, where essentially you have some determination of where mine activity is located globally, and then you can derive an electricity mix from there.

But, our most current data is, based on our last update of what we call a mining map, which is where we work together with mining pools. And this is from January, 2022, which means that, , quite outdated, right? Specifically with a very [00:20:00] large share of Kazakhstan and still a very large share of China in there.

And then you get a choose mix that might not be perfectly, reflect how the reality looks . This was why it was so important to, even though having a very large participation of US firms, but still seeing trends activity coming to South America activity, coming to the Middle East, activity to some extent, coming to Africa, to other nations , in Asia, right, for instance, Bhutan, these trends that are relevant.

We still saw continuous activity in Northern Europe specifically. So I think this was very important to have this often anecdotally, , refer trends. That activity is taking place now a lot more in these regions, basically, I would say validated , by the findings in the report.

And naturally those things have heavily influenced electricity mix, right? Because specifically looking at Kazakhstan, looking at China, the share of specific coal, even if we look at, [00:21:00] at fossil fuels, right? There’s a huge difference terms of emissions per kilowatt hour between coal and gas.

That it’s really, it’s double or even more than double. So it’s a huge difference what, even if we have fossil fuel mix, right? what specifically? How the fossil fuel mix, the composition of the fossil fuel essentially. , And if you take everything together, we just see that the emission intensity is quite lower than what we currently estimate based on.

Mining map update and, and I think this is one of the key findings that the report really helps. So while we are very confident in terms of electricity consumption, estimates that we’re ongoing, we provide on an ongoing basis, it’s really this translation from electricity consumption to emissions that is this difficulty where you need a lot of data that is extremely hard to get as I’m sure you’re aware, right?

Because , there is no technical way to determine where, , folks are located. There is no improved stake. For instance, it’s a bit [00:22:00] easier where you can develop network crawlers and these things determine.

Running, let’s say a Bitcoin full. No. And where actually taking place, because if you use a map of full nodes, most likely you will have, I dunno, 40% in EU countries. Or something that is Germany. Most likely it was 20% or so. If you just look at Bitcoin full nodes. And so it’s really important to know specifically where mining activity is taking place.

And , that’s , , very important just for this emissions calculation. I think this is what the report complement our more theoretical work.

JohnPaul: And there’s no way, to your point, without going out in the field and really asking these large miners \ and medium sized and small miners. Where they’re located, how much hash rate they have, what’s their power mix to collect that type of data? Because the binder doesn’t say, Hey, I’m mining from Argentina.

I found a block in Uruguay. It doesn’t say that. It’s very hard to, differentiate the zeros and ones, and [00:23:00] you guys have done a great job highlighting that. I’m really doing the legwork to give us some institutional level research in this space.

Alexander: Yeah, thanks a lot. I think this is where we see ourselves having this, great, , I would say connection to the industry, but also, to more of the, investors, regulators, policy makers, these stakeholders. So I think this brought access to stakeholders.

We have from, , , working with you guys to source the data to then provide something that is , , , universally useful. I think we can really function here as the bridge specifically between, let’s say, academia and practice , and assuming this role as a trust data aggregate, right?

So, for instance, if you a minor, and please feel free if.

, I, I will reach out and ask, , do you wanna, your, have your logo displayed or don’t you have a logo displayed, , if default, we just, , collect the data and everything is anonymized. But if you say, Hey, [00:24:00] well, , we wanna be shown, as contributors, that is perfectly fine as well.

So really having this trust in the community as well for someone, because I think that’s the big point, right? Where you need to convince a miner to give you data because, , there is, just the, upside is quite limited. Only if there is an upside, if a lot doing it right , to actually help us to get them really research out there is impactful.

This might be the upside, but kind the difficulty getting in and much appreciate,

JohnPaul: and to your point, it, it cannot be done without these minors. So everyone who submitted data, thank you. And if you haven’t submitted data or if you’d to participate in the next survey, make sure you follow up on the website and there’s QR code links in the report on how to sign up. one of the questions I have for you here is how should policy makers, which are these stakeholders that you engage with balance?

Bitcoin mining’s energy demand with innovation, especially in regions with fragile grids or maybe stricter ESG mandates, the [00:25:00] eus MICA framework, how does this affect mining and , how should policy makers or and people regulating the grids be viewing Bitcoin mining as a resource?

Alexander: Yes. What we are trying to do is, being, not having an opinion in that regard. , so what we are not trying , to take a position, , what we’re trying to do is to inform, and I think hopefully this comes also across in the report that we are highlighting specific, we are not recommending, let’s say, any grid operates or policymaker to, okay, this is kind situation, this is what we’re trying to recommend you to do.

, It’s more trying to put the information out , and if we get a specific questions, obviously we’re gonna answer this question. And, I think there are use cases that are being described that might, in many jurisdictions, in many specific circumstances, be very helpful. But it’s then basically , for those grid operators or policy makers to, decipher themselves, how to view it.

We’re not actively trying to, [00:26:00] influence any policy making that regard, but we are putting out the information, right? It’s out there and those obviously who are interested or, , interest groups can pick it up or those stakeholders can pick it up and read it , and then say, well, this is something that might be helpful to us or it’s not something we need, , whatnot.

So, yeah, unfortunately I can’t give you a specific answer for that. We really trying to keep our impartiality here. What we are really trying to do is aggregate information, put out the information, and then, let the public take over the conversation.

JohnPaul: No, but I love that because it provides, you guys are a neutral, force and neutral data aggregator that then these policy makers can look at as someone without an incentive to, let’s say, , make Bitcoin mining or you don’t have a financial incentive to, to grow it. You’re more just reporting on the facts and that I think that’s a key to, moving this conversation forward.

How did you guys come up with the case study? So you guys interviewed different groups, you came up with a question, , the report says, for example, beyond [00:27:00] base load, can digital mining help support power grids? And then you answered that question, you might have a couple charts in a case study. Can you talk a little bit more about maybe how those were conducted?

And, you mentioned the natural gas opportunity. Maybe if there was other case studies that highlighted something that you were not expecting or didn’t know before you started, this journey.

Alexander: So happy to see a couple of words on the case studies. So the case studies are all based on public information. So that’s basically, the case studies were used to, I would say. Make things, make explanations more tangible, right? Because sometimes you’re gonna get lost in descriptions , in these things, right?

That are a bit more abstract. So after explaining certain concepts, these inside boxes and case study boxes could help to essentially make this entire concept that is being described or show a real world use case for this concept. And, , I hope they’re [00:28:00] helpful, but it all is based on public information.

So we did not interviews or anything in that regard. And yeah, so I hope that they’re basically just there to really make those descriptions. For instance, what you said the topic beyond base law, which essentially demanded response, right? We have ther case study for, mitigation.

We case as well for with the heat project , in Finland, right, where they utilize waste, heat and things. We yeah, have some inside boxes that should also help make certain concepts a bit tangible, to the audience. This is why we.

JohnPaul: I definitely think they do PR, paint a real world picture of, to these point these miners where you can point of actually where they’re doing some amazing work. One of the items you guys touched on was a level of concern for selected challenges amongst Bitcoin mining firms and the top three were long-term energy price increases, [00:29:00] unfavorable government action and adverse development in PTC price.

Can you talk more about how you guys investigated those concerns? Or did you just in the survey, ask them? They rated them and then we didn’t actually investigate maybe how these concerns could affect mining as a whole.

Alexander: Yeah, , so basically that’s based on a survey question. So, essentially, those answers were given, so naturally we did some investigation right to came up with them in the first place, with some pain points, which I. Frankly, I think, yeah, for, for those who follow, this might be particularly interesting for those who might not on a day-to-day basis follow the industry, but for those who follow the industry and, , starting, , with this very volatile environment in China when mining was still there, to other countries having their opinions on mining, right?

This all has implications to some extent that, , , are you allowed to Any mining there, which naturally is.

Those things are [00:30:00] really important and we see it now even with Russia. Right. Specifically , I think more this south and ian region where they first initiated temporary bans and now I think fully banned until over two 30 or

Marker

Alexander: something. so those things matter. We see it, , yeah, in some areas in the US it’s, it’s very interesting, right?

Because you have this very different opinions so there’s not a US opinion essentially besides obviously the administration’s opinion, but depending on the state, you might face very different opinions. , So to understand the regulatory concern, energy prices, I think it’s also pretty clear because .

We did the survey as well about when it comes to cash based expenses, about 80%. So that’s your major expense, right? And if you are not certain about long term electricity deals or prices , that you are able to obtain, it’s very hard to model your profitability going forward.

So, yeah, , there is, I think , lot of backstory. We did, a bit of description in the report as well about this, but, basically we did not, ,, too deep [00:31:00] investigations. I think , I might put in a couple of examples, , but yeah, basically all those were obtained just per minus, essentially rating them.

JohnPaul: And one of the things you, You also touched on was historical trends in hash price across the different type of having events. Is there anything that piqued your interest in either Bitcoin’s price or hash price or machine efficiency in those macro metrics that all come down to the imports into the profitability of mining when you were doing the report?

Alexander: I think mining economics is a super interesting and really fascinating topic, right? specifically you have a lot of different variables and as a minor, naturally you need to do forecasting, specifically , where you expect bit the big comprise to go. Where do we expect hash rate to go?

If we think about the. In US dollars. There’s a lot of , with the Bitcoin price you have, naturally the increase in hedge rate where you have all these components. If you just look at hash spread in Bitcoin, , it’s a bit simpler, right? Because you just know it’s going [00:32:00] down, most likely, given just the halvings, right?

The point is, I think always , in this conversations, the topic of transaction fees, it’s just that besides some temporary anomalies, so this spikes right? We saw where we really, really saw a, , we just need to go back to the halving. , that was, if you, I was actually observing a lot of developments specifically during,

Marker

Alexander: because I saw this fascinating, , the amount of revenue derived from transaction fees, right?

I think it was the minus future dream of having transaction or already transaction fees being in BTC terms and multiple of the actual block subsidy. , So, but yeah, this was as, as we saw, really just temporary. , if we look now, we’re, sometimes it’s, , zero and, , one set per, we buy in these things.

So there’s a lot of interesting component there as well. And what I conversation is currently I. Mining firms to [00:33:00] continue. But as we go along, halving by halving, the question will really be , , how this dynamic will change, right? Will transaction fees pick up , or what will, essentially happen, right.

, Naturally, , you can always say, well, Bitcoin is going to, I don’t know, 800,000, 1 million or whatnot. Whether you believe, , this is obviously, , people have their own beliefs about this. , , it’s just that it needs to continuously go up, right? To an extent that , it covers halvings and it covers increases in hash rate.

And I think one of this, and this is really my personal opinion, one of these points, I believe the mining ecosystem at some point will change. And we’ll, when it comes to consumption, we will peak at some point. Where. Simply miners will need to be a lot more flexible. We see it already, some miners starting to, for instance, utilize excess renewables, right?

, specifically if you have the rere where , you’re not mining a hundred percent uptime, you’re mining maybe 30, 40% uptime and really trying to capture, , excess energy when it’s pretty [00:34:00] much , we not call it for free, but very, very low cost. And possibly we’re gonna see this developments more as, most likely, hash price will be in, , my opinion, , as we move forward, as we move forward in time, will become a lot more volatile than it’s now simply because, , sometimes lot transactions, which might significantly impact.

The overall block reward, right? If you have a very low block subsidy, and sometimes you might not have that much and miners might actually adjust, their activity depending on the network activity at any given moment. This, but this is again, , my personal thoughts on this. So I think mine economics is al already very important, but, it’ll be become much more important, , , as we, , move along each having and, , yeah, generally I think there’s a lot of, I don’t, there’s a very specific conversation that I’m not trying to avoid and to make things , not too controversial in your podcast, but I think it will be very interesting to see this.

JohnPaul: And one of the things about hash price that you’re are hinting at is [00:35:00] this evolution of a Bitcoin miner from being the most efficient machine on the market, using a hundred percent. Of the power to maybe only using 40% of the time having energy to it. You guys really dug into e-waste and the evolution of the mining machine.

What came out of that from that, research and was there anything that shocked you or what were you seeing about a migration of mining of Bitcoin miners? Are they migrating from the US to other places? Can you touch on that report and on the report when it comes to e-waste specifically?

Alexander: Yes, absolutely. I think this was a very, very, , interesting finding, frankly, because there was not really information out there right. Specifically about, . Sim we did a simplified calculation about e-waste, to say this as well, but I think the key message that we, what we found there is that, , what is it, 88, close to 87% is being repurposed, resold, or recycled, which I think , if you look into computing, the e-waste of the [00:36:00] industry is, to me, seemingly definitely a variable that should be considered to some extent.

It would’ve been interesting and possibly for next survey to differentiate a bit more between reselling, repurposing, and actually recycling. , But overall, basically the amount of e-waste, which is about what we found about certain percent of phased out hash rate, is definitely something to consider if you do, I would say any e-waste modeling.

But in the past we really didn’t have much information that regard. Right. There was not really where well. Some miners, some public miners put something on their website. So we do recycling, we do this and that. But really something on a more network level where you have engagement , of private miners as well.

And to get some understanding of, okay, , what does end up , as really as e-waste, , a certain amount of phase out hardware. So I think that was really an an interesting finding and hopefully something that, spurs further research, outside, let’s say our survey because a lot of [00:37:00] researchers, , who might want to look into this again, and, trying to, , refine estimates or new estimates in,

, Point about distribution of rates. So, as previously mentioned. I think definitely something that needs to be said is that, we had a lot of engagement of US firms, right? And I think this really comes through, if you look at the distribution of hash rate, where we found the US at 75%, which, personally I think is overstated, by any means, it’s, the US is def, I have no doubt that the US is the world largest mining hub, but again, where we are exactly, we have possibly on the lower end, around 40%, but then we’re at the higher end, , 75.

So it really depends where we’re , then. But overall, I think, , given that the US is a very large country, you have, access to all kinds of different energy sources, right? It’s not a country that predominantly powered by a specific energy source, [00:38:00] for instance, when we. And, so I think overall our findings , are representative, but it needs to be said that, , there might be a bit of a US overhang of participation, which you think, which still, right? we have 50% of the network cash rate. So it’s I think still a very, very interesting insights. But yeah, when it comes to the specific geographical distribution question, I think it’s more about trends.

So where we say, well, if you look into the current map and compare it to our 2022 estimate, we can see that, well the US is still the num number one mining hub right there. There wasn’t any change. But what we see now is specifically activity in South America. We see activity in the Middle East. We see still continuous activity, \ to some extent in Northern Europe.

We have some countries Bhutan there as well, which have been anecdotally, referred to, right. We see some emerging activity in Africa. So I think it’s more about this trends that, are often. Not so much Africa, [00:39:00] frankly, I hope that will change in the next report because yeah, lot has happened since then actually.

So, , but still having specifically South America in there having, , the Middle East emerging and, , so it’s more about direction development, , developments that happened, let say since 2022, where this is helpful, but I would not, I would not, know, say well there is this specific percentages , are, , something that is, , absolutely true where we have also very low share of Russia, for instance.

Where we know that there is definitely a lot of my activity in Russia. , China is really a black box, frankly. It’s, , getting any data from there, , obviously you ask, but, , it’s, it’s very hard to get any data. There are estimates out there that have some data for China.

I have no idea where they get it from, frankly. , And, but, no one really knows. But there is definitely some , of aspect that I hope that in the next survey definitely trying , to increase global outreach and keen to, to have more , , non non-US firms participating as well.

But still, , even [00:40:00] if it’s a bit , , the sample is a bit us focused. I think overall, , most of the findings besides essentially the global distribution are still very much relevant.

JohnPaul: And I definitely agree with you on that. And one of the things that you mentioned was , you’re getting all this research, you’re not trying to make an opinion about it. How hard is it to be an evidence-based researcher versus speculating and having these, opinions about all these different topics when you’re putting this report together and what goes into.

The hours, the team, the coordination behind a report this, how long did it take? Can you talk more about the process of how, actually, how this piece of information came about? I.

Alexander: Yes, absolutely. I would say, let’s start with, I think the most important, the absolute foundation of the report is trying to get a lot of stakeholders together to provide you with in.

Marker

Alexander: This is really where things start. And a lot of work goes into reaching out, engaging, and [00:41:00] trying just to get, just trying to get responses right, that there is, at least I would say 35, 40% just goes into this.

And, the rest is then essentially the, the write up. But yeah, it’s, there’s a significant, significant time. You really need to do engagement. You need to get, , folks to contribute to convince them that, hey, this is actually, it makes sense. You might not have an immediate benefit by spending, the whole 30 minutes, one hour, depending on , if you already know questions.

or if you need to dig a bit deeper into your company, that it’s worth to spend the time. Right. And I edition.

It might, it is definitely something that people are interested in, so it might make sense for me to contribute. And I think for the first edition, this was the most difficult part because it hasn’t been done before, at least not , for mining specifically , , at this scale, right.

We had the mining council, but it was four or five questions. So it was not that extensive. It’s a lot easier to ask to answer those questions. [00:42:00] 25 questions that in some cases go a lot deeper than what was asked there. , So, and then the first iteration was really the difficulty to convince, folks that, hey, well, it makes sense to contribute.

I hope that we will have, with the first edition done, , a good starting point where we can convince more that say, Hey, well, if you had privacy concerns, , look, if you wanna be stated, , you’re gonna have your logo there. If you’re not, you’re not right. There is no , specific individual data sharing that we are doing.

, And I really hope this helps. But yeah, this is, I would say, one of the main power we started with this, I think about the last year. And in April, first of all, you need to, , develop all the survey questions and these things so that, that makes sense as well. So we had some, , conversations with, , some key players , of the former, of the BMC, , to give, give you just feedback where, , you have a bit of back and forth , bouncing off ideas, what is possible and what is not possible, what is too granular and how far we can still go.

Lot of this , thought process as well. And the [00:43:00] end, if you in the end have collected data, you need to do the write up, right? This is depending, this time it took also longer than expected because yeah, in the end it was okay, let’s create something that really enables, , someone who might not be perfectly familiar with the subject.

Two, starting at page one, ending at page, I think. And you might not be, an absolute mining expert, but , what Bitcoin is. , , on the surface how mining works, , the current of the industry, about trends and you get a bit of finance aspects as well.

If you read the appendix, right? So you generally mining economics as well. So you have a, a general understanding, okay, what is actually happening? And we have this condensed into this one report. and yeah, this is why it took a bit longer, but yeah, it wasn’t really nearly, nearly a year essentially that we were working on this combined.

So there’s a lot of, lot of effort that goes into creating something this.

JohnPaul: And now you have a great piece, which you can build on top of to your point, and you have that core introduction and some of these questions on pools and in [00:44:00] firmware. Can you talk a little bit more about how you got to the pools in firmware?

, So as you were digging into this report and creating it, you guys also talked about firmware and mining pools. Did you end up interviewing any pool operators or firmware operators, or was that data collected from the miners, themselves?

Alexander: Yes. So, basically the mining pools, I took it from man space. I think they do , and shout out to them. They do fantastic work. So when I look about anything related to, more, this, more specifics about just looking at blocks, right? Block compositions, things, it’s pretty much man space.

Otherwise we work with, which is fantastic by specifically collecting data on, so yeah, we did not interview anyone, , in that regard. Specifically one of the. So the report. So naturally we had some conversations and, and really appreciated his input specifically for some nuances, in the writing in terms of the firm, where yet that was a survey question.

So that was basically [00:45:00] about asking Myas, okay. Essentially what firmware you’re using. And I think findings are quite interesting. specifically we have, what was anything that unexpected is the large share of just stock manufacturer firm, where ish was something that I did not expect.

Frankly, I, I know that there people were, were using it, but, , it was more than one fourth, I think, more than a quarter were using it. So , that was actually one of those things that was surprising. But yeah, in proprietary firmware, right. Specifically, it makes sense if you’re really large mining company to invest it in and build your own software, right.

Specifically tailored to what you need in your operations. , And yeah, and you have some , other operators as well, but yeah, I think it is, similar to the ASIC market where it was also a, i I would say not too much of a surprise with, , Bitmain being heavily dominant and then, , if the other two manufacturers, micro PT and Cannan, this will I think be something where, on a time series will be quite interesting to, observe, right.

So specifically having now or [00:46:00] see, see how things right. June, 2024. But if we, , if we end up at some point at the third iteration, how things changed over time, right. Specifically how market shares changed in terms of, , the ASIC market. Because so far, I think this is also where we’re contributing. It was known, when you do researcher, you need to actually data point.

It’s one thing to, , , have anecdotal evidence or know things because that, that’s generally known as an instrument practitioner. Then having actually, okay, this is a data point based on, , primary data from minor. So that’s specifically if you do any research, it’s just really valuable to have this data point there.

and yeah, very, very interesting to see how these things evolve specifically with, , most likely now US manufacturers entering the stage. And also in terms of firmware, right? How things change there as well. If we see , third party firmware usage increasing, or if we continue seeing still this quite, I think [00:47:00] about 40%.

So, yeah, I think data are.

JohnPaul: And I think that is where you get a, a really good set of data over time where you can start telling the story of the industry, which, it, it will be amazing to have this data set. So, as I said, thanks again for putting this together. One of the last things is you end the report with this idea about data center infrastructure, ai, HPC and Bitcoin mining.

was that something that you think miners, were they saying they’re looking into it or they’re actively exploring it? It seemed it was a small portion of miners. Do you expect that to be a larger portion of the future? So when it comes to AI and data centers, how are you seeing Bitcoin miners transition into AI in, in the report?

Alexander: Yeah, sure. So happy to elaborate a bit on this. So that was essentially, we had only a very small question in there trying to, understand a bit to what extent [00:48:00] participants. Mine Bitcoin, to what extent they mine other cryptocurrencies and to what extent they, for instance, have power allocated to AI operations.

And I think this is one of these data points that we will very closely observe as, , we progress, for instance, addition by addition, because that I think, is a very interesting data point that you can go back historically and then compare to, for instance, new information as, , we have, , in the end a more theoretical chapter, which is not based on survey data, but basically just us thinking about, well, there are the synergies with AI in many ways of, , established mining companies, right?

Having, having power contracts, having the already existing possibly fitting infrastructure, having that talent. To, diversify their business. And this was one of those things in the sentiment questions that we also saw where business diversification was one of the key areas that miners were looking at.

And, and also [00:49:00] if you look into the really statements of public companies, I think miners moving to AI is, is something that is definitely very, very interesting component. And, , yeah, we, we did observe this. I think it is given specifically when we think of, the way how most likely the block reward will move.

Having a lot more volatility possibly when it comes to revenues derived from mining activity, having this more, I would say, less volatile income stream, , providing AI HPC services. It’s quite interesting for mining companies. but on the other hand, what we see is, and I think what we found that the amount of CapEx required to build this infrastructure system on a whole different level.

So, it is for, I think, , in, in terms of the, the capital requirements to enter this space. It’s, yeah, it’s quite a feed for mining companies, right. given that if you wanna build a hyperscaler, it, it’s gonna be very, very expensive. And, yeah. So [00:50:00] there are a, a lot of different angles.

And specifically also trying to understand the, specifically the increase in demand for, or the increase in demand of, for instance, state of the art. . Training models, the computational requirements and how those have changed over time and these, so I think overall there is, , there is a lot of information in there that might help understand , okay, this might be an interesting field for miners and explained why specifically having established power contract, possibly some infrastructure that can, that is already existing or , can be retrofitted on.

And having that, the talent pool that allows to leverage those synergies and yeah, but still on the other hand, there is this CapEx element that makes it not so easily translatable where say, Hey, well we have M Bitcoin and now we’re just doing ai. We’re actually okay.

Fantastic. Yeah. So basically one of the key differences between. Building mining [00:51:00] infrastructure with actually building AI infrastructure is the CapEx requirement where we show that essentially if you include GPUs as well, it’s about 33 times. But even if you just build the infrastructure, it’s already, I think about, about eight times.

So, while this might be very interesting field specifically with this, , very much increasing demand, we also have a couple of charts there that show where this demand is coming from, but much more computationally intensive, , trading models, but naturally, I think mostly from inference and, we see a lot of miners looking into this now, and it will be very interesting to see specifically how the power location changes.

Because currently the snapshot we have from June, 2024 is I think a very good starting point and will be just very interesting to see how, this power location changes. Right? , most likely it’s expected to increase. But I said before, you can have a lot of megawatts of Bitcoin mining activity or generally cryptocurrency mining that, might not one-to-one translate [00:52:00] into, okay, we build one megawatt of, bitcoin mining, let’s say, and one megawatt of, , AI compute.

Where in terms of the cost it’s just not comparable. so, but it will be quite interesting to see how things evolve. And we see a lot of public companies looking into this and we have for those interested, there’s a lot of information there in terms of electricity consumption, these estimates how much overall data center power usage is expected to be in 2030.

so it’s a, I think there’s a lot of, nuggets of, I wanna call it nuggets of wisdom, but it’s a lot of in interesting information, I think , in the last chapter of the report specifically, about, where we are most likely heading in terms of mining. , You mentioned some point.

, Move towards the transaction fee based model from currently predominantly reliant on block subsidy and naturally having this synergies in terms of infrastructure with AI makes, , and diversification on the business model being one of those points we found out in the survey to be specifically [00:53:00] interesting to firms.

, I think this will become more and more important and definitely something I want to monitor even in much greater detail going forward because simply, , I think , it’s fair to say that, , the computational demands of AI will just significantly increase. I just remembered a tweet from Sam Altman when we had this entire studio, Ghibli Studio Ghibli thing coming up and he said, yeah, our GPU are melting.

, which I can imagine, right? So you You need to have in the future a lot of computational power. , and I think that a lot of mining firms are very well positioned to captures opportunity becoming a more diversified firms where we possibly might also see firms that say, Hey, well we want, we wanna be a, a pure play.

That’s just our main business model and we’re gonna want to keep it that. so it’ll be very interesting to observe this different pathway minus take.

JohnPaul: Well, Alexandra, let’s close it up by talking about what’s the reaction been to the public, you [00:54:00] know, what type of news outlets have reached out to you? Any cool appearances that you wanna highlight, , to the audience and , how are you feeling after a year worth of work and really being able to see a come to fruition?

Alexander: Yeah. It’s so. It’s very great to see that there’s a lot of public interest, right? We see a lot of, , , first of all, I think, we got interviewed by traditional media, Fox News. We got naturally featured in crypto native media. There is a lot of new things coming up, where we had, I think a lot of, , folks were token 2049 and now coming back and see the new information that is being out there.

So we had a couple of requests in that regard as well to, , have a, a podcast, have a, , , some interview or doing some other things. So there’s a lot of, lot of public interest. I think this data in many cases where really highly anticipated. Then this overall very well received.

Naturally there is a bit of that, this criticism around still, even though we have captured nearly half [00:55:00] of the network hash rate, that there is still a bit of the US focus, but that’s what it is, right? , in the end, , it’s, I think a very good starting point, but there’s a lot, lot of things I think we can, we can do better as well and, and trying to engage more people, more geographically diverse.

But overall, I think what we managed to do is providing a deep insights into this. Very much, , some, some very nitty gritty data points that usually just don’t have, right? So I think we, we managed to provide a lot of, , valuable insights in that regard and hopefully create something that we can build on and, and improve over time.

So we can really manage to capture, , at some point maybe 65, 70% of the network and having it geographically diverse. So really creating something where we’re essentially, as close to the ground truth as possible, let’s say that, requires an extraordinary amount of engagement from a lot of different stakeholders.

And I hope that with this report we, we set just the foundation to do so. Personally, [00:56:00] yes, I’m, I’m really happy because, , there’s a lot of work coming together and a lot of time spent on this. And, , yeah, also happy that it has been published, right? Because, , , you wanna also move on to your, to your next project at some point and then have some , , bit of time for yourself as well.

But yeah, we are very, I hope that, there will be second edition and, , we can, we can continue this work because I think it’s just something that is specifically for those who are fam even, even for those familiar with the, with the industry, but also particularly for those who are not familiar with the industry, to having this , , data set that, , , provides this insights that you otherwise just don’t get.

JohnPaul: And some of those datas dataset that you highlighted are 138 terawatt hours. Annual consumption of Bitcoin mining, 0.54% of global electricity, 0.08% of global GHG emissions with 38.9 metric tons of carbon dioxide. , 52% of electricity from sustainable [00:57:00] sources. 82% market share held by Bitmain, which you, you, you highlighted us being the primary hub, 75% emerging markets of South America and the Middle East, which you highlighted. What is, if anything, other key highlights that just , come to mind as we wrap it up here or anything you wanna leave the listener with as they read the report? Something they should really focus on or really make sure they can dive into?

Alexander: I think it depends on the person interest, right? So I don’t wanna highlight anything in particular because the point was really to have this very vast data set. For me personally, one of the key points were to compare our, in terms of electricity consumption, to compare our theoretical model estimates.

At a specific point in time, which was the time of the snapshot, 30 June, 2024, with, , those electricity consumption estimates that I can derive of from the survey data. And then given these are entirely different methodologies, and then see [00:58:00] whether there is a very star discrepancy, which would indicate that, well, , possibly our theoretical model, something, something isn’t work right, or there’s a huge difference between practitioner insights and our theoretical modeling.

but what, I saw was that there is very little difference actually. So it, it, it feels that , that, our theoretical work in terms of electricity consumption has been corroborated, which is quite important because, , these reports are not issued on a daily basis.

So, , things change quickly. We just from June last year, right? , hash rate skyrocketed frankly. So currently it might look very different, right? And, and so it’s important to have. For instance, our daily updated estimate there that can tell you and, and , that it’s reliable about, okay, what is the current state of today in terms of electricity consumption and, , until the next report is available.

So I think this was, for me, personally, one of the very important points. Also, having this, , new insight into the electricity mix, I [00:59:00] think is very important because our past estimate is really based on very outdated data, so it’s important to have here an update as well. And, , naturally, , everything is important.

Right. But I think particularly interesting as well was the e-waste issue where we saw that okay, repurposing, , reselling, recycling is just a huge component. And yeah, a lot of things as well, specifically when we talk about, , the amount of, , power coming from otherwise flat gas. And there’s a, I think a lot of, a lot of interesting information in there.

but yeah, this is my, my insight, but I think everything is important and it really depends if you read the report, if you’re interested in a specific area, but I think we covered so much ground that, , whatever you’re gonna look at, something might be in there for you.

JohnPaul: And we will be linking the report in show notes, as well as the, , online UpToDate theoretical model that you’re mentioning. So feel free to look at those to learn more. Alexander, thank you so much for coming on and remember [01:00:00] to mine on.