Transcripts are autogenerated. May contain typos.

00:00

welcome back to the mining pod today we're joined by Fred Thiel CEO of marathon digital this podcast was produced in person at the recent Jackson Hole ski Summit so there's some Ambiance and background noise in the show thanks to Amanda Cavallari for help in making these shows possible today we discuss Marathon digital's Playbook including its relationship with applied digital after the compute North bankruptcy movement into the UAE s9txp purchases an increasing minor efficiency under its hosted model

00:26

[Music] [Music] thank you Fred welcome to the mining pod we are in warm sunny Jackson Hole just kidding we're inside in the basement of a hostel recording this got a little Ambiance going on which is great we're going to talk about Marathon digital uh one of the largest publicly listed Bitcoin miners out there a pleasure to have you on the show today it's great to be here awesome well you just finished up a panel and a lot of it was about strategy unfortunately as we said before a recording here I did not get to attend

01:03

because I was doing another podcast but we can go through here now talk about strategy um we are way down Marathon to produce its Q4 numbers which I wish we could get them on this podcast but another day we'll get them uh looking forward to those numbers but let's talk about strategy and and talk about some recent headlines so to kick off the conversation you guys are now moving towards mining in the Middle East you guys have worked with applied digital to secure hosting in the US you're continuing your guidance for 23 extra

01:31

hash in the US continuing that hosting model uh you've sold a little Bitcoin recently but mostly modeling secondly largest huddle of any public company to my knowledge so a lot of talking points there but all that derives from a strategy that you and your team are working on I'd love to hear just a high level explanation of that and then we can get into like the individual bits sure so we've spent the past um year kind of setting up our execution for this year uh which is obviously building out and deploying the full 23x

02:00

hash by mid this year and we're well on our way um February production report shows us that we make great progress there and March will be a great month as well with a lot more progress there we have a lot of minors racked and ready to energize and the energization is now starting so we're feeling very bullish about that yeah we're very focused on energizing and optimizing and part of optimizing is also taking the full technology stack and similar to the Apple iPhone in the Apple ecosystem Apple can do very

02:30

unique things with their phones because they not only own access to the Asics but they own the operating system in the phone they own part of the stuff at the carrier level they own the whole ecosystem around apps Etc and so we've taken a technology approach to our strategy and said you know if we're going to really be very successful here we need to be able to control our miners from top to bottom and so what does that mean well we run our own pool and by operating our own pool it means we don't have to essentially optimize to

03:03

the lowest common denominator across a bunch of different miners with different fleets we have a pretty homogeneous Fleet it's either x19j Pros or x19 XPS and by mid this year our Fleet will be 60 or hash rate will be 66 percent XP so 30 more efficient than the J Pros uh and the most efficient miners commercially available today if you look at our Fleet composition overall our efficiency you know across the board will be about an average of 25 joules per Tera hash where the average mining Fleet out there today

03:37

is in the 45 range so very efficient Fleet um in that regard so we also with the learnings from kind of the third party hosting model have figured out that we want to have more and more control over the actual operation and so we spent the last year developing a strategy around owned and operated more vertically up integrated and the deal we've done in the UAE is the first kind of manifestation of that we designed that full system you know we're obviously controlling the build on it and defining how it's going to be

04:11

operated and again because we control the whole stack now uh you know we have our own firmware running in the miners we have our own controllers running miners we have our own pool running we have full end-to-end control and eventually control all the way down to the Asic and what that does is lets us really tweak and optimize production the UAE site is an immersion site we ran the pilot site for that and it was 110 days before an engineer had to go do anything at the site and let's just say we overclocked XPS a

04:44

lot faster than anybody ever expected and I'm not going to give you the number but it was a lot faster than anybody expected and so when you combine the fact that an engineer didn't have to touch anything and we can manage it totally remotely because of the software stack that we've built you know we had the benefit of you know 2022 was a tough year but we spent that year investing in technology and we're not seeing the fruition of that so that site in UAE it'll be 250 megawatts when it's fully

05:10

built the largest data center in the Middle East um and so we're super excited about that and you know that'll be online by the end of this year it's an 80 20 joint venture so we don't get the full benefit of that hash rate we get 20 of the benefit but that's additional to the 23x hash we're deploying here in the US so we have about 7x a hash running at King Mountain in West Texas we've since turned on about an additional little over 2x a hash in the month of February on top of that we have

05:40

a lot more coming online with applied digital uh in March through July which we're super excited about and all the applied digital hash rate is XPS so we're super excited to see those start really contributing to uh to our hash rate but as we look forward you know our model is more really about us partnering with energy companies and then designing building and operating sites ourselves and uh you know there are a lot of opportunities out there today to find access to power there are a lot of people who built sites either to do

06:14

third party hosting offer third-party hosting or for self-mining who don't have the capital to build it out and so you've got a Transformer you've got substation you've got some basic infrastructure and they just haven't done any more so we'll look at some of those types of opportunities and leveraging the design we built for UAE and replicating that domestically and we think again with the success we've had at the pilot site in the Middle East if we can replicate that in the U.S like in

06:42

Texas and we manage our power pricing properly with hedging Etc we think it's going to be a huge game changer for us so we're super excited about the balance of this year and you know into the having you know a follow-on logical question is well what are we going to do beyond that 23x hash yeah so um you know we'll have to see we're adding yeah we're tripling our capacity by mid-year compared to where we were at the end of last year the having happens in May of next year it's really depending on what happens

07:13

with the price of Bitcoin right if Bitcoin stays under 40K by the having that's kind of the equivalent of 20K today they're gonna be a lot of miners that are going to have a struggling time uh they'll struggle in 2024 beyond the having and typically bitcoin price if you follow historical patterns doesn't really start ramping until the few months after the having and then tends to Peak kind of 6 to 12 months after the having yeah um not sure that that'll replicate but if bitcoin price remains subdued and uh

07:44

you know there's not a lot of free excess Capital floating around the world right now so there's a high likelihood that may happen we believe there'll be a lot of consolidation opportunities come next year and so we've been very focused on building a lot of cash on our balance sheet uh deleveraging our debt and putting ourselves in a position where we have lots of optionality which if you've heard me speak before optionality is kind of how I like to run a business and so we'll have the ability to either go

08:08

buy machines that are cheap go buy sites uh acquire somebody if it makes sense haven't yet come across a deal that makes sense because you know one of the unique Dynamics in this business is that machine prices get cheaper the lower bitcoin price becomes and so if you're going to acquire somebody you could just go buy new machines for the same price so lots of different things we could pull on from that I want to go back to the hosting model which for those listening are where the compass runs this podcast and has also been admire

08:36

with uh the hosting model hosting model is pushed to its limits for the last two years uh large deployments lots of different contracts in people's hands and those contracts have not always play on people's favors compute North is probably the poster channel of that I mean we don't really need to go into detail there too much but now that they're in chapter 11 uh and all their contracts have been bought at auction from other people the question becomes should people be pursuing that model again or should you

09:05

just build your own infrastructure I tend to agree that like hosting can can continue just about building correct relationships uh but you guys seem to agree with that as well apply digital some of you guys work or you've uh extensively especially North Dakota want to get your thoughts on the hosting model so yeah that's just like the the just the question where is hosting for you guys how do you guys view it and how does it view how is it viewed in your strategy for 2023 2024 sure so um as I was talking about the strategy you know

09:35

when we design and operate our own sites that kind of obviates the need for having a third party do it now unless somebody can do it better than we can and we're more than happy to learn from people and partner with people who can do it better but as long as we think we can do it better than what we're really interested in is people who are interested in developing sites and then we'll partner with them to develop a site and we'll own build and operate our own part of that site but there's a whole power component that

10:05

you know we don't do obviously um I think the third party hosting model works for miners of a certain scale right it you know it takes a lot of capital to build a site it takes a lot of capital to operate a site and um you know if you're not talking about tens of extra hash you know the third party hosting model is still very attractive now I think it's changing we're before if you go back to like 2021 it was all about fixed price deals and it was kind of you had a fixed price on the energy and there was a

10:35

fixed markup for the hosting and that was your nut and you know that model got companies that core into trouble because they had variable pricing on their input cost of the energy but they're they're selling at a fixed price to their customers and so a lot of the third party hosters now are saying hey you know we'll basically charge you a fixed fee per machine for hosting operations and then it's passed through Power and either you can let you know the hosting company do the hedging and do all that

11:03

or you can do it yourself at that point all the risk is on the minor it's not on the hosting partner the only risk that's on the hosting partner is the raising the capital to build the site and then uh you know obviously operating it so I think the model's changing and that's creating even more of an incentive for companies of e to the medium scale to start operating their own sites now it starts coming down to can you get power pricing that's appropriate you know attractive can you hedge do you have the

11:33

financial were referral to do that and not all companies have the balance sheet to do that we're lucky in that we have a very strong balance sheet and can do that but so that's I think an area where the folks that are in the third party hosting business meaning the service providers uh it's going to be a tough couple of years for them you know um Jason lust was on the panel with me today here and you know their model with Corsicana is kind of you know a blend of self-mining and third-party hosting and

12:00

the question is you know there's not a lot of demand for the third party hosting so I think uh you know companies who follow that traditional model uh are going to have some challenges and they'll just have to do more self-mining which yeah I think is good for the industry at the end of the day but yeah I think the third party hosting model you know it can work small sites smaller volumes of miners but the bigger you get the more you need to vertically integrate yeah not to mention the lawsuits that spring up from third

12:26

party hosting uh yeah we won't mention any of those but they are there's a lot in the industry right now let's talk about the other side of that facility model and talk about the play in the UAE um with adgm you said that that is to be past that 23 exit hash guidance which is exciting at the same time A lot of people are looking you guys sort of being the first big Mega Miner to step into the Middle East to my knowledge at least and they're like oh this is kind of far out of a limb um what do you guys think

12:56

about that in terms of like what was the impetus or motivation to move out of the United States out of North America really and start mining uh in a very different jurisdiction with different political uh ideologies and political ways of doing things different Power structures contracts there's so much knowledge you guys have to build up again just to be able to operate there so yeah great question so um yeah Bitcoin miners always focus on finding the lowest cost energy and consistent Supply in the Gulf region you have a huge

13:28

climate asymmetry right you have Winters that are cool and you have Summers that are extremely hot and so you have to have power generation enough to support all the air conditioning you need in the summertime which in the case of UAE is somewhere around four gigawatts of energy and in the case of their winter time they only use about one gigawatt so you've got this huge swing so there's a lot of stranded energy if you've got another Dynamic there which is they need the heat off take from their energy

13:55

generation to run their water desalination because they don't have wells they have to desalinate water from the Gulf and turn it into fresh water to drinking water so they still have to run part of that infrastructure to generate heat and so in addition to that uh you know in some countries there the government subsidizes electricity to their consumers so they're kind of you know getting it they have issues on all sides of it so um the ideal play kind of in that region is if you can take that excess energy

14:28

that's stranded energy and consume it generate some revenue for the energy producer so they can maybe decrease their subsidies or help fund more water desal whatever they're going to do with it then it's a very symbiotic relationship the advantage you have in that region is often the government controls energy generation energy distribution they also control the rules and laws could be risky as well the benefit with UAE is they have this construct around the adgm which is essentially the Abu Dhabi Global markets which is a

15:02

international trade zone that operates under English law not local UAE law and so foreign companies like Marathon have the benefit of having a more a set of law that's going to uh for commercial purposes that is more easily understandable by both our lawyers and by courts and um it's a very unique thing and this is only unique in UAE which makes it very attractive it doesn't really exist in other Gulf States and at the same time obviously um you know the UAE is a very wealthy Nation they have a sovereign wealth

15:37

funds and and Deep Pockets which to invest and so it was the ideal place to kind of try and find a partnership with um an entity that was willing to co-invest with us and yeah we're a minority partner so it's really us providing technical expertise and knowledge and being able to build this site and it's been a great experience so far for us and we're really excited about you know the site coming online later this year so this is going to be somewhat of an ignorant question because I'm not aware of that

16:04

Region's power Market too much but I would assume that it's mostly natural gases powering a lot of these applications is that more or less correct or yeah so the traditional source of energy in Abu Dhabi has been natural gas but they recently commissioned a five gigawatt nuclear plant which is green energy and as you may or may not be aware the I believe it's the president of UAE is the chair for the cop 28.

16:32

for whatever the time period that lasts it's a year or two years or whatever so they're doing a big uh the next cup 28 conference will be in UAE and so um they're very focused on green energy and so essentially their whole yeah the government in UAE is very focused on doing everything uh around green energy they want to do you know Green steel all sorts of green things so essentially you know we will either fully offset um any fossil fuel component from this uh or it'll be sourced from nuclear one of the two and now is a reason or a

17:09

motivational factor to move into that region was the green energy so that was exactly because you know our model is you know we've stated we want to be carbon neutral what we're doing and you know that's why we cite behind the meter at like wind farms in Texas and North Dakota um it's our focus is really trying to be as much behind the meter renewable and in this case it's uh you know it'll either be nuclear or fully offset yeah gotcha gotcha okay let's move over to s19xps and talk about that again uh as

17:35

you mentioned earlier you guys are moving towards XPS for a stated reason which is the efficiency gains uh the tidbit about the overclocking that's news to me most people I've been talking to have had a very difficult time cracking into those machines um so if we can get more information I'd be Keen to get that out of you um but yeah I would be curious to get more thoughts on the satanx B series and the reason for purchasing them because some people are looking at it and saying like the initial Capital you're deploying to get

18:04

those machines is not worth the efficiency gains over a stated period and that's notably Zach from clean spark is one who says like hey don't get the XPS we're gonna hold off they have purchased some to be fair but uh some people pushing back on that a little bit because it's just the initial capital is high what's your thoughts on the my economics with these well it's all dependent on what your cost of energy is right if you're paying more than four cents a kilowatt hour come to

18:28

having s19j Pros aren't going to be very profitable so you know you can either decide to not mine yep or mine at a loss or you can mine at least be making some money and you know what a lot of people don't think about who are not in the industry is it isn't about what's happening in the next year it's we're constantly looking at kind of four to five year deployment Cycles So when you buy Miners and you deploy them you're basically either Contracting for four or five years of hosting and you're expecting

18:58

four or five years of economic life out of the minor and so in that four or five years you'll go through at least one having at least one winter and at least one boom period and so if you are only profitable from the spring through the Autumn and the rest of the year you're unprofitable you're never going to get a payback on those miners no matter how inexpensive they were and you know when we bought the XPS we placed the order at the peak of the market but we priced an order with price protection so you know we didn't pay 85

19:30

a Terra hash for them in the end uh with a lot less than that obviously so um you know we believe that you always have to be at the bleeding edge of the efficiency curve uh especially when it's not a number go up market and you know we're definitely not in a number go up environment at least not for the next few years uh same topic slightly different immersion we've been waiting on Argo and Riot to give us some information about their immersion setups but information from various bars you guys have a state intention to dip or

20:00

dunk a lot of these Asics and get more immersion benefits out of them any updates on how that's been going so far given that the immersion Market's so difficult to work in there's so many different people trying to build these things and it's been been fairly difficult sure so like I mentioned earlier you know the UAE site is immersion it's single phase and you know the pilot ran without human intervention for over 100 days and with significant overclock and so why can we do that well again we

20:26

control the full Tech stack from the pool all the way down to the firmware in the miner and the controller in the miner um if you were to ask me the follow-on question which is where do we think immersion's going yeah we're big Believers uh in the fact that single phase immersion is good but first gen single phase immersion is not efficient you have to use second or third gen and what we've deployed in the UAE is second and will be third gen as well over the life of that because the initial deployment will likely be Gen 2

20:56

and then by the time we do the balance it'll be gen 3 on that immersion Tech but we believe that dual phase immersion is where it's really going to be longer term benefit with dual phase immersion while today can cost a little bit more from a capex perspective it allows you even better hands-off operation of Miners And I think the bigger change whether it's single phase or dual phase is we're going to go from shoebox miners to blade-based miners which are built for immersion you know the problem in an immersion tank today

21:29

is because of the physical space the metal enclosures of miners take there's a lot of wasted volume right so if you instead went to a blade server model with a black back a black Bad Lab sorry a back plane where you have blades plugged in you can get up to four times the density of hash rate in the same physical space and if you can do that properly then now you can start building really compact data centers and if you can if you can operate those very hands off you now don't have to work at 100 megawatt scale you can go down to five

22:05

megawatt scale and when you go down to five megawatt scale you can start doing things like methane flare gas you can start using landfill you can start using small Hydro sites you can start using true stranded energy that has no place to go and that's a whole different ball game yeah and when that technology trickles down to the medium and small miners to buy then all of a sudden they're actually at a competitive advantage to the big Miners and so I think that will reinvigorate the kind of mom and pop and

22:39

the small scale miners because you'll be able to go and contract for a flare at a oil field somewhere get a Generac and then deploy one of these you know immersion blade-based systems and run it and go out and check it once every two weeks and you know otherwise you're getting Telemetry over your your iot system definitely well while we're on the subject of efficiency gains I want to talk about the the stack that you guys use and this is somewhat returning to early part of the conversation but same with efficiency

23:07

here you guys often uh are working with hosting providers and then flying out texts if need be but for the most part you're leaning on the hosting providers to provide service the issue with that is often that that person doesn't have incentives aligned or they don't have the training must be loved out with this at this point how do you guys approach that um seems to be a tech Solution on one hand but noting that Tech can't solve everything how do you guys get to a point where you have the same or greater efficiencies than

23:37

other uh Bitcoin miners in the space noting that you guys already have like very leading efficiencies so one of the things we do is we work very closely uh so for example in our King Mountain site where now that usbtc is take your HUD 8 usbtc whatever the name is going to be going for I think it's new Hut I think it's an or hot Corp how to eat Corp um so we work very closely closely with the usbtc team and kind of showing them here's how we want you to run the site uh they have a lot of great skills and

24:10

power management energy purchasing and stuff like that curtailment technology so it's this very symbiotic relationship where we're learning from them they're learning from us yeah um you know we have uh built our own repair center centralized repair center we deploy remote repair centers at all the sites so we're able to service our own miners uh more efficiently and you know especially with the uh in the relationship with applied digital there we've had more influence on the design of some of the of how those miners are

24:40

uh operate and going forward you know we'll obviously have a lot more control over that so again it's kind of for us we're growing out of the third party hosting model uh as we've kind of reach adulthood if you would uh and uh taking the training wheels off if you would but uh we're super excited about the future and and what we're doing and yeah we love how the industry is maturing and um you know I think the the challenge going forward for a lot of my nurses is going to be more about just having to be

25:09

much more sophisticated and how they manage finances and manage you know their growth and it's not just number go up or you know and grow at all costs because you know that model just doesn't work while speaking of usbtc and had a uh you guys are working with usbtc at that King Mountain site and Huddy and I'm just merged hot date is naturally competitor as also a public minor do you see any potential issues there or any bottlenecks in a potential merger between the two companies uh with Marathon or how do you see that evolving

25:40

no I mean I joke with Jamie Levitt and that have eight who I've known for a few years and respect very much uh that I'm not your customer so you better be nice to me for once um but no I think it's a very collegial relationship I mean I have very good relationships with the guys at core the guys that Riot you know everybody I think we all have to this is an industry where you know while we're all competing for for you know our fair share of the global hash rate yeah um we're also all fighting external forces like regulatory

26:09

environments energy policy things like that and we all have to work together and be able to work together and there's definitely um synergies in in you know leveraging the best of uh breed in different companies for all miners to succeed so love that okay last question for you as we close out here what are some forward-looking things you guys are looking to do or things that you're going to see in the mining Market this year and Prelude to that question is 2022 you guys invest in the tech and now you guys are reaping the benefits of

26:41

that from 2023 into 2024 where are some things that you guys are thinking about from a strategy perspective um so as I mentioned a little earlier I think the price of Bitcoin is going to drive a lot of the decisions so we kind of operate um three plans at all times bear model base case bold case right and there are three very distinct plans and we maintain the optionality um to be able to execute on that so what does that mean well a we have to have the balance sheet that lets us react depending on what's going on in a bear

27:10

Market case uh you know right now our goal is just to get deployed um so we can Harvest between now and the having as much as we possibly can the question is you know will the will machine prices um where will machine prices be between now and the end of the year based on the price of Bitcoin because that'll drive 2024 growth it also uh you know you also have to have a look at kind of where Energy prices are going so you know we have these very complex models we have strategy team that are constantly kind of crunching trying to figure out

27:44

exactly what strategy is ranking highest of the three options uh at any given day but I think essentially and this is kind of Beyond 2024. I think you're going to see us um dabble in um other types of renewable energy than just solar and wind but you'll see us partnering more and more with developers of solar and wind or owners of sites again continuing the behind the meter philosophy but really working on joint economics when you partner in the ownership of sites um it gives you a couple of advantages uh the most interesting of which is you

28:26

get to share in the incentives the government pays these companies to build these sites and our single biggest input cost is energy so if you can find a way to subsidize your energy cost you have a competitive Advantage because it means you can mine when other people can't if the price of Bitcoin drops and we focus more on being able to operate in the downside cases because the upside's easy just grow yeah that's an easy strategy it's in the downside that it's hard and so we spend a lot of time

28:56

analyzing and thinking about how do we protect ourselves on the downside and is there a way to get to zero cost energy because if energy doesn't cost you anything you can mine and grow as much as you want yeah yeah awesome well Fred thank you so much for joining the mining pod appreciate your insights and hopefully speak again with you soon thank you really appreciate it

‍

‍

‍

‍