TFTC: A Bitcoin Podcast - #399: Bitcoin Mining is Igniting an Engineering Renaissance with Rete and Isaac from Cathedra

Episode Date: March 2, 2023

Join Marty as he sits down with Rete Browning and Isaac Fithian from Cathedra to discuss the origins of off-grid bitcoin mining, their findings on over/underclocking ASICs, and the direction the mining industry is headed. Check out Cathedra: https://www.cathedra.com/ 7:38 - Great American Mining: Origins 18:24 - The state of off grid mining 22:17 - Designing for diverse domains 24:01 - Deadly details 31:09 - Pioneer challenges 34:57 - Applying bitcoin and mining everywhere 39:52 - Cathedra research on over/underclocking 48:29 - How heat affects chips 53:34 - Finding the sweet spot 56:47 - The future of mining and energy 1:12:55 - Changing the ASIC market 1:19:40 - Bullish vibes, creating an engineering renaissance 1:31:27 - Wrapping up Shoutout to our sponsors: Unchained Capital River CrowdHealth Bitcoin Talent Co TFTC Merch is Available: Shop Now Join the TFTC Movement: Main YT Channel Clips YT Channel Website Twitter Instagram Follow Marty Bent: Twitter Newsletter Podcast

did you mute my mic it's on now i can hear myself now logan logan i'm sorry for keeping you waiting it's been 10 000 years since we recorded the episode two hours two hours same thing 10 000 years in bitcoin time 399 399 rip 399 I sat down some good friends of mine

I've been working with for quite some time in the mining industry Reed Browning and Isaac Fithian two killers that not many people know about I think you guys are going to enjoy this one

especially if you're in the mining industry if you're looking for some alpha on running mining operations we just leaked a lot of it did any of the alpha get on you Logan it

evaporated in the two hours it took for the episode the alpha has evaporated into Logan's skin this rip was brought to you by our good friends at River River is a bitcoin company built by bitcoiners for bitcoiners this is personified not personified

it's implemented in their product stack they have an exchange buy bitcoin if you dca into bitcoin using river you're not going to pay any fees if you custody bitcoin on river you can have certainty or peace of mind knowing that your bitcoin is in a multi-sig wallet with 100 reserves backing it river doesn't leverage any third parties uh to to buy bitcoin for them they've built their own infrastructure they've built their own wallets they've built their own libraries they've built their own lightning nodes

you can send and receive bitcoin over lightning on river they have lightning service river lightning services excuse me which is an API that allows you to build apps on the lightning network if you're a developer they have mining capabilities

want to buy an ASIC and plug it in with a hosting contract river setting that up for you what else do i have to mention yeah it's bitcoin company built by bitcoiners for bitcoiners they're at the cutting edge of bitcoin technology always have been and this is because they're an old bitcoin only company with focus on bitcoin allows them to do all the cool things that you want them to do you may have your exchange already but have you tried river give it a try go to river.com

slash tftc use the referral code and after you buy certain amounts of bitcoin you will get some free bitcoin i don't know if i'm allowed to say free bitcoin you'll get some bitcoin bitcoin river.com slash tftc this trip was brought to you by our good friends at unchained capital as well they're right down the hall and they're here to help you eliminate single points of failure in your custody model with their vault product is two or three multi-sig wallet where you hold two keys unchained holds one they have a consultation team that'll get you set up if you get unchained.com slash consultation set up a call with them they'll get you set up they'll get you comfortable they're really leaning into retirement right now if you're a bitcoiner out there

just thinking oh gosh what am i going to do with my bitcoin if when i die or when i retire inheritance not retirement why am i saying when you retire to the afterlife also known as inheritance they have an inheritance protocol that they built out for you if you reach out to their consultation team they can teach you about this inheritance protocol too something to think about sooner rather than later. You don't want to be the Bitcoiner that unfortunately passes away and your family doesn't know how to access your keys. Unchained has created a product and a protocol for you to solve that problem. So reach out to them, unchained.com slash consultation. A lot of retirement plans, retirement to the afterlife plan. That's

what we're going to go with. This trip was also brought to you by good friends at CrowdHealth. are here to help you reimagine how you pay for your healthcare. It's not health insurance. Health insurance is notoriously opaque, expensive, impersonal. They don't really care about you. It's really hard to get in touch with people. CrowdHealth is doing it a different way. It's not health insurance. You pay a monthly fee that gets built up in a dedicated account. If you ever have a health event, you get your bill, you give it to CrowdHealth, they negotiate your price is lower with the doctor. The doctor is willing to do this because they get paid in cash right away. You put your bill out to the crowd health community. You pay the first $500

and the rest gets crowdfunded by that community. A hundred percent of the bills have been paid to date. Can't guarantee that moving forward, but the model seems to be working. And on top of that, you're joining a community of healthy individuals. So the overall healthcare cost of the community is lower there are some uh there are some metrics you have to hit to get into the community you have to be relatively healthy the bmi of the crowd health community is significantly lower than the national average so if you want to reimagine health care pay cheaper rates i was on cobra on crowd health now paying is obscene i was paying an obscene amount on cobra it is significantly cheaper with crowd health i'm extremely happy with it go to joincrowdhealth.com slash tftc

sign up using that code tftc just plug it in last but not least this trip was brought to you by bitcoin talent co bitcoin talent co is here to solve your problem in the space which is the jobs market connecting the talent that understands bitcoin with the companies that are looking for that talent there are many bitcoiners out there many of you i'm sure who are listening who work in incumbent tech firms on wall street in the banking industry and you love bitcoin you've around for a while but you don't really know anybody in the industry but you know that you could provide value to the industry go to bitcointalent.co connect with the bitcoin talent co team because they are here to provide that bridge for you to the bitcoin industry likewise

if you're a company looking for this type of talent reach out to bitcoin talent co as well they are going to create a a service that really fine-tunes for your needs they're going to help get potential uh talent up to speed with what your company does exactly what you're looking for whether or not they're a good fit they're going to work with you to fill your needs at the talent layer layer we'll just go with layer use uh go to bitcoin talent.co if you're talent looking to get a job put in your resume if you're a company looking to find talent uh put in your information tell them the tftc sent you enjoy this rip with reet and isaac you've had a dynamic where money's become freer than free

if you talk about a fed just gone nuts all all the central banks going nuts so it's all acting like safe haven i believe that in a world where central bankers are tripping over themselves to devalue their currency bitcoin wins in the world of fiat currencies bitcoin is the victor i mean that's part of the bull case for bitcoin if you're not paying attention you probably should be probably should be we're gonna isaac we're live sweet is this your first podcast second it is yes well i i do some uh some podcasts on the side with friends that we uh we it's called the disagreeables and we uh sort of uh disagree well as sort of the play on words there but other than that yes nothing nothing formal nothing bitcoin specific so this will be my first bitcoin podcast

marina i can recall one you did with tom a few years ago oh dude that that was a a while ago and i've taken a hiatus ever since but you know it's been a while it has been and i'm happy to bring you out of the hiatus to bring you away from disagreeables to tftc isaac oh love it freaks how to preface this i mean obviously if you've been listening to the show a while you know that i've been involved in mining oh yeah um not you too i'm talking to the audience right now uh they know i've been into mining but i'm really uh the ugly face in front of the mining operation and everything i've been involved with with in the mining space whether it be great american mining or cathedra and some side projects uh these two

gentlemen i'm speaking with today are the brains behind the operation actually plugging in the machines setting up the management systems really designing a process and farm management and what else, logistics to make everything work. I think you guys are two of the unvaunted individuals in the space who have done a lot of work that many people don't really know about yet. So I'm excited to get you guys on to highlight what you've done because you guys have done some incredible stuff. I think we should start with Great American Mining,

reet particularly like talking about that first pilot container that we had up in utah oh yeah the other one yeah you could barely call it a container uh but but yes it did mine some bitcoin so that's what had that going for it uh yeah so i mean reet and i we met back in 2018 we were working on a hydrogen project together this was before great american mining uh it's like a pure R&D phase. And, you know, we reached a point in 2018, I mean, this is the depths of the bear market when, you know, we were trying to integrate mining tech with this hydrogen tech and it ended up, you know, just not working out. We decided to move in a different direction. So we're demobbing this site, pulling out the miners that we were using. Yeah, the Dragon Mint T1s is what we had

on site. So RIP, anyone who bought Dragon Mint T1s. But yeah, basically we're pulling all these miners out and put them into storage. And I was just like, gosh, I wish there was something we could do with these things. And so I was chatting about it with Reet and he was driving home one day, looked at this giant flare out in Utah. He's like, wait a minute, there's a bunch of energy there. Maybe that we could do something with this. And so this is before, um, before Reid and I were formerly working together, we were from two different companies working on this R and D project. And so, um, I, I went to, um, you know, my, my boss and it was like, Hey, look, I think we can do this. He's like, all right, look, I'll put a bunch of money into this. It hasn't worked

out. Uh, you know, I mean, drag him at T ones. I mean, come on. I can't blame him. And he's like, all right look guys i'll let you do this crazy thing i've never heard of anything like this you got five weeks and you still have to demob the site and so reet's like how the heck are we gonna get gas in five weeks like like this is just a just crazy task we got it in three days you know it's like reet look you're our guy you know if anybody's yeah yeah well that was all I had nothing to do with that. And so he's like, he calls up his buddy at a wastewater or a water processing facility, you know, from the produced water that comes up when you're pumping oil. He's like, hey, you know, can you just give us some of this, this gas that's entrained in your water that's coming in? He's like, oh, it sounds crazy, I guess. We'll try it out. And so he was like, all right, sweet. We got a place to do it.

now we just got to go so like literally in the morning we'd go to this site this r&d place i would be taking out miners you know electrical panels wire you know we're taking apart the site right i'd be taking stuff out loading up carrying it over to and oh this is the other bit reet had a buddy out there and he's like hey can you just give us a a patch on this gravel driveway uh that you've got just so we can just you know build this mining thing he's like uh i guess sure and so literally i'd load up in the morning take the the parts over to his friend's driveway and we would like unload it and start you know and we'd be building we had a like a hand crank generator for power there's no internet so yeah that was uh that was kind of the origin of genesis

yeah and every now and then we look at me like this is this is crazy like like why are we even doing this look it's just like uh it's like like the first iron man movie we're in the cave we got a box of scraps and we it doesn't need to be pretty it just needs to get us out of the cave and then we can build better stuff and we did it did it in five weeks and it was enough to justify building a second one one that we actually had a budget for we weren't just using spare parts and yeah that turned into great american mining yep that is a great american mining origin story freaks reed what were some of like what were some of the main lessons we learned on that first proof of concept deployment

well if you think mining is sexy it's not it's hard you get bloody knuckles you get everything else and you don't know where you don't know what questions to ask if you haven't done it before um because all the iterations that we went through on those containers like the second one was our air intakes on the first one were like that big then the second one it was the entire wall you know then bigger fans bigger you know just everything got scaled up uh because we found problems in airflow and then we realized that the first christmas tree design was a horrible one because it took three hours to change up the filters whereas you know a different orientation took 10 minutes so everything was you know you you can put everything on on paper and it's it's

going to look okay but when it comes down to actual functionality it's like you have to have user experience you have to have um you know the right engineering that goes into it you have to have um especially if it's going in remote sites functionality that you can do from your couch right and all that has to layer on top of each other yeah and so and that from experience yeah and to that point i was gonna say it's at that point doing stuff from your couch that includes building remote systems that allow you to to tap into the the site and control fans control um the the airflow walls like if it's cold that christmas tree design maybe it was hard to change the filters it did keep snow out correct yep uh yeah after two or three uh

two or three repairs yes it did yes like yeah i mean oh sorry go ahead no you go ahead yeah Yeah. I mean, really, one of the things that I've sort of come to is you can solve these problems with one of two ways. You can do it with people or you can do it with technology. So, you know, in general, you can think of mining facilities as falling into largely one of two camps. Either you have a campus style Bitcoin mining center where you've got, you know, everything is in one central location. You have, you know, big, beefy power lines bringing in power site, or, you know, you've got one, you know, sort of hut where, you know, people camp out. You can do your repairs out of there, and, you know, if something goes wrong, you send a guy down to whatever aisle it is, and they go and fix it. Then you have sort of these more remote style builds, and that's what we were doing at Great American Mining.

you know you've got a pad two hours away from from the nearest airport and if you got to go out and do maintenance on that it it's tough to get out to and so we we had to have you know command and control systems where you could adjust the temperature um i mean the very first bills we did nothing was automated we hadn't figured that out yet and uh you know uh props to austin on that he did a really great job taking uh some of those long nights that that we did at first and uh you know automating like okay you know this is how the miners work best what are you guys doing when let's automate that so um you know you did a great great job you know really really helping us um you know scale up great american mining and yeah it was

just uh you know you're not going out to a site easily so you need to have good information so that you know what's broken if something breaks or you know what the problems are um because you got to figure that all out remotely because if you're driving out there you've got to have the tools you need, the parts you need, you know, whatever it is before you're, you're going out there. It's really time intensive. So, you know, remote control and good information are key if you're going to have a distributed style build.

Yeah. So what would you guys, I mean, obviously upstream data was around Crusoe was around when we were building the first iterations of the great american mining boxes uh and more and more players have come to market since then you have giga a bunch of other players i can't even name off right now but uh what what would you guys say is the state of the market of off-grid mining particularly with these containerized systems and the control systems is it night and day from where we were in 2019 2020 or are people still learning on the go here everybody has gotten better i have to say like even gift props to steve usually when i look at

a design i can say okay i know what he did and why he chose to do it that way if it's a good design, right? Like everything is a trade-off. There's no such thing as a perfect design. It's saying, what are my hierarchy of needs? And they're clearly defined downwards. These remote boxes are not the cheapest boxes on the market. So, you know, right now cost for most people is at the highest point in their needs, like, oh, I can get a box cheap. But that doesn't mean that it's going to function remotely in the way that you want it to,

or even in a campus style, for that matter. So if cost is at the highest and you need to have more automation or you need to have more data, that's not going to be your cheapest option. So you're going to trade off, as Isaac said, the people component is going to be a bigger portion of your operating costs at that point. Yeah. And so what do you guys think is the Goldilocks perfect trade-off model for these off-site boxes or off-grid boxes? excuse me well perfect i mean you have to have automation uh like on off on the the um out each

outlet right like that's that's number one but you know our new design has like plc's kind of lots and lots of variability with it i think you can get away with kind of like microcontrollers and get something a lot more scaled down and still make it work so you know cost is going to be how can you get the same amount of automation for the lowest cost and and really hone in on all the data that you really need and that's kind of where the the step is right now saying all right how do i pare this down to what i need and what resources can i use because a lot of these boxes that that especially like steve even digital shovel they have heat recirculation for cold areas right like that's that's something that we've done since the

beginning it's like we need to recirculate some of the heat if it's a super cold area and bring that into the intake so that we're not blasting the miners with like negative 20 degree negative 40 degree air so all the the all the manufacturers who kind of listen to their customers have said okay here are the problems that we've seen and then they start adding on some of those those functions and that has that has been very apparent over the past couple of years yeah you touch on a very good point there like not all off-grid mining is the same obviously If you're mining in the Bakken up in North Dakota, your, your needs, your pressing needs are going to be different than if you're off grid mining

in West Texas. And how do you design a system for that? Yeah. It's really hard to make a one size fits all system. Period. Because in Texas, you're probably, you're bare minimum. You're going to need to have like a water curtain and adiabatic cooling in North Dakota. You don't need that. period. But you need a lot of heat recirculation. You need to have kind of environmental protection measures. The biggest thing, the biggest enemy is really dust for a lot of these miners as well.

So environmental control, the minute you put in filters, you're going to have pressure drop, you're going to have resistance to airflow. But you're protecting your investment, you're protecting your miners so is it really worth it yeah in my mind yeah yeah so i i mean reit brings up you know a spot on point which is that you know every site is unique you know in particular if you're doing uh flared gas mining you need to get a gas analysis sometimes you need to treat that gas there's a litany of of tiny details that come crawling out of the woodwork anytime you say oh i just want to mine here it's like okay well you know what we would look at things like what's the prevailing wind direction you know what i mean like we would just reach

really good on that sort of thing once we sort of identify hey we want to do this he just goes he's like an electron microscope he goes right down and we'll find the smallest of things that are he didn't know were important it's like oh wow you know i didn't think about that uh you know you could uh you could call them an electron microscope or an ant fucker like yeah there it is there it is yeah what is the ant fucker yeah what is the definition of an ant fucker read in your mind well let me put it this way i don't consider myself an expert but i definitely know what has caused me pain and so i i would consider it as all right that's caused me a lot of grief how do i fix that and how do i bring that and make that next iteration better so that that's

why i would consider an ant fucker it's like all the things that may have kind of led up to this point of where you want to focus your energy because it has either caused you a late night or freezing cold temperatures or just time really yeah so the answer the very small details so i guess what you're getting at is the very small details and these off-grid operations are very important yeah yep i mean you know one thing for example is when a gas well or the gas from a becomes unstable you don't want to have your gen sets running at full tilt right so you know do we look at the generator and see oh is it going to shut down or do you look at the gas and the gas supply pressure and so like kind of figuring out where you need to look on the the

energy supply system that's where you can get a competitive advantage because you need the containers or the miners or the load on the containers to pare down shut off and go into self-protection mode because if they don't and you don't have eyes on it you know we've seen the the horror stories of of you know caked on snow into machines because the door wasn't closed you know you don't want that a very small and actionable detail that everybody should close your fucking doors people the snow or the dust getting in yeah but i mean you touch on something else there like we've been talking predominantly about like working within the box and the whole mining operation,

but you mentioned the whole other side of that, which is the power generation gensets, which is like a whole other Pandora's box of problems and intricacies to work through. Your primary goal is to keep the genset up and running. Mining is actually your tertiary thing. So you want to make sure you have gas, your gensets stay up,

and then your miners are on, right? Because without that cascading sequence, you're not mining. Right. The big thing is, you know, if you don't think about how these systems integrate with one another, you're going to pay for your lack of knowledge or lack of planning with downtime. I mean, you need to have an understanding of how is this well working, how do the generators work, how does that all tie into the container, it all matters right and you know for instance if you turn miners on too quickly on an off

off-grid site you can you know basically cause the generator to suck down gas faster than the pipeline can sometimes provide it and then your whole site gets shut down and then you're sending technicians out there to get things back up and running again it becomes a mess pretty quick and so you can automate some of these things but you know really since each site is sort of unique in terms of the systems you're tying into, you need to do an analysis of, well, how does this system work now? And is our current build a good fit? And going back to your question about, you know, hey, what's the sort of the Goldilocks zone? I think probably, so in general, I think the space is still developing a lot. I don't know that anyone has sort of cracked the code of, hey,

this is a one size fits all build. And part of that is there's a litany of little details you have to think about. I think probably what will end up happening if I was sort of a, you know, try and pull out my crystal ball and say, hey, I think this will happen. I think you'll probably see some sort of a base model that suits most needs, and it will be built in such a way that you can easily have these bolt-on solutions. So, you know, regardless, you know, you have to have rack that the miner sits on. You have to have wires that bring power to that miner. There are certain things that don't change too much. And I can see a world where people, you know, where somebody figures out, hey, this is the base model. This is the bare bones, what we need, you know,

in most markets. And here are these simple add-ons. You can quantify exactly how much they'll cost and say, okay, if we're going to go down to West Texas, here's, you know, here's a water curtain or, you know, a water mister bolt-on, you know, with maybe, you know, some kind of water collection on the roofer or something, right? And then you go up to North Dakota and all of a sudden, you know, heat recirculation becomes more important. So you sort of create these different bolt-on, you know, technologies, you know, and also, you know, in terms of, you know, remote control of power cycling, I think the industry is

going to head toward, okay, here's your, you know, think about like, you know, an ant box where it's just you know bare bones basic electrical okay well let's get something that would allow us to do automation as a bolt-on technology so if you don't need automation because you're in a campus style you always have have it staffed you don't need it but then if you want to do a remote deployment here's this bolt-on package you can just add it to an existing container and then you're off to the races i i think that's where we'll head making all of that work well is is a really challenging task because there's so much going on that it's not as simple as you might think once you actually start getting into the details of it but i i think

that's where we'll end up heading so the uh the market for purchasing containers is going to be similar to the market for buying a car where you decide what what gadgets and upgrades you want like i think exactly walk into the dealership so to speak what does it feel like for you two to be on the cutting edge of this budding industry like tinkering with this stuff some of the first two individuals i mean i imagine the population of people working on the particular problems that you guys have worked on is relatively small yeah i mean in the beginning it was it was nuts i mean There were no YouTube videos about this stuff. It was just like, okay, you know, here's this sort of electrical video on YouTube that sort of says something about what we need, you know, looking up articles on HVAC, you know, thermodynamics, you know, textbooks from way back when. Yeah. Yeah. I mean, it was it was pretty crazy because there weren't like good resources. It was just written. I know sometimes we felt like, you know, dumb and dumber. I'm telling you, like, because we're just sitting there like, I don't know if this is crazy or if maybe this will work. And we just had to try it to find out.

Um, you know, but it was, but it was always a case of, look, we just need to make enough right decisions. And, you know, part of that is just staying really humble. And, you know, just when you're in the thick of it, when you're, you're in the pain, say, okay, you know, reality is going to tell me, uh, what it, what it needs to make this work. And I just need to listen and stay humble and just work hard. Uh, you know, sweat, sweat equity will go a long way if you're willing to stay humble. And I think a lot of, a lot of other miners, a lot of other people in the space, they're in that process and, you know, people are putting in the sweat equity and, you know, they're going to, they're going to make it. I mean,

that's what we had to do. It was just work hard, stay humble and, and, you know, be alert. You know, the information's all over the place if you know where to look. Yeah. Listening to what, what the data is showing you so you you have to kind of overbuild to gather that data and kind of i mean we didn't know i mean one of the mistakes that i made was i sized the fans incorrectly because i didn't think about the expansion of air on the hot aisle by 11 percent right so my like the fans couldn't keep up the the hotter air expands in volume and you know it's like i'm just not quite there yet and and so everything after that was you know had a at least a 20 to 30 margin so right it's crazy to think of i mean you guys mentioned earlier a lot

of people think like oh it's simple you find a strand of gas well you find a generator you plug it in it works but this is again a cutting edge engineering problem that a lot of people are trying to solve right now like when that was like the most fascinating part about working in great american mining um working with you austin tom now you guys continuing that at cathedral like i've learned more about like physics uh petroleum engineering uh electrical engineering you're telling me in the last five years than i ever would have expected like if you told me i would have learned this uh about this particular domain when i was 21 i would have told you you're crazy but that's what bitcoin mining particularly in this off-grid environment forces you to do

but it's energy right like it's it's how do we get to energy and and that's really how i got into the bitcoin space is like there's such this beautiful relationship between energy and and bitcoins i just see a lot of waste that goes on in the energy space that it it's uh as much as i hate saying is like bitcoin fixes this right like energy has this this why waste it when you can use it for something and it's it's not a scarcity thing it's saying how can we be more efficient in the way that we use it and and there's no better way well at a fundamental level uh just just one of my overarching principles in thinking about this stuff is you know look at look at nature you know some of humanity's best inventions comes from

from mimicking what what nature is doing and uh you know what does what does nature do with energy and if you look every waste product of of an animal or a plant any organism on this planet any waste product is food for something else that that's just that's just how nature works at a fundamental level and so you say okay what is bitcoin food that it just you know and it's very rudimentary but at a fundamental first principle level bitcoin requires energy you know specifically electrical energy and it's like okay well i need electrical energy that doesn't you know, just, you know, pop out naturally in, in the world. So we need to find other energy forms, but still at that fundamental first principle level, you say, okay, where is there wasted

energy? What, where is that the waste product? And, and that's where Bitcoin can industrial scale. You can do that with flare mining at a very small scale. You can use Bitcoin miners to heat your home, right? Because that's something that you need anyway. And, you know, I mean, in New England, we just burn oil in an oil burner to heat homes. You know, we don't have, usually don't have gas hookups. You know, worst case you have, you know, just there's literally people buying, you know, space heaters, just electrical space heaters. And I just prefer mine to be subsidized by Bitcoin. But, you know, the point is, you know, there's wasted energy all over the place. I mean, I just think about clothes dryers, like just tumble dryers. Tumble dryers in

the US use more power than the Bitcoin network currently. I mean, what if you just dried all your clothes with Bitcoin miners, just integrate the chips into a dryer? I mean, anywhere you see just energy being expended for its own sake, Bitcoin can insert itself there. And that's, it sort of gets at what we, our mantra at Cathedral, hey, we like to be the cockroaches with the cockroach model because there's all these little nooks and crannies where energy is being wasted and Bitcoin can thrive there.

So it's just sort of the ethos of it. So we're not crazy to think that Bitcoin fixes this. Is Bitcoin going to make us as energy efficient as humanly possible? Yes. If we're not in the effort. Yes. I mean, the use of energy

is what allows humanity to thrive. And if you want humanity to thrive, it means you need more energy. I mean, people use heat for lots of different things, and that's the primary waste product of Bitcoin. So, you know, again, that fundamental principle, okay, what's the waste product? How can that be a primary input for another task? And Bitcoin can thrive there. So, yeah, Bitcoin fixes this.

Gets me all bullish. Why don't people realize this yet? you know that's a good question because it's out of the norm it's out of the box i mean i've kept my hot tub warm all winter using waste heat from a miner just one i subfloor heating gets preheated by a bitcoin miner like do i pay for the electricity yeah Well, I also am experiencing, you know, zero degree weather. So I need heat too. Yeah.

So, I mean, off-grid Bitcoin flare gas mining is just another step removed from the electron. It's just one step above that, so to speak. So it's just finding where the forms of energy that's useful and how to harness that. It's a no-brainer. Yeah, it is a no-brainer. We have some asymmetric information here, which brings us to the topic why I really brought you guys on right now.

We've been focusing on, like, the actual design of a mining operation, but you guys go even deeper than that, and you've been doing a lot of research on underclocking and overclocking particular ASICs. He posted a big thread from the Cathedral Twitter account a few weeks ago. They got hot, and a lot of people seemed to love it. But I think to highlight the breadth of knowledge

that you guys have in this space, we just described designing and architecting these environments in which the miners live, but then once the miners are plugged in and running and you have a stable environment, there's more that you can do at the machine level that you guys have been tinkering with so obviously you've been tinkering for years now um what what have you guys found you know

the bear market gives you an incentive to change your perspective i i do have to say that You know, everybody looks at a big, you know, when you say aftermarket firmware, everybody says automatically things overclock. Oh, I'm just going to push farther and farther and farther. It's not necessarily that. So, you know, and Isaac helped me kind of see this. when you start conventionally when you say hey i'm making a container you are physically restricting the number of and i call them parking spots but rack space and saying i have x many rack spaces for x many machines and i need to then size my electrical to that and say okay this is how

much power this container is going to use. But what if it was a flexible thing? So it was saying I am restricted. What you're saying is I am restricted by my rack space. Well, right now, the dollar per terahash of machines is cheaper. I'd rather buy a second machine at a more efficient and run it more efficiently because that's going to increase my operating profit, period. So the way we were looking at and sizing things physically was it's a static number. But what this aftermarket firmware has been able to do is that it's given us a knob, kind of a fourth dimension knob to say, well, what if it isn't? What if you had more rack space? If you already have it, then why not utilize it? Especially if

you're going to get more efficiency out of it. If you've already bought machines and just don't have enough power to plug it in, but you have rack space, why not utilize it? So it's, you know, in the cockroach mentality, how can I maximize all the assets that I have, whether that's power capacity, whether that's rack space, whether that's, you know, restricted by power cost, Whatever it is, you just kind of look at all those variables and say, is it really something that's static? And it shouldn't be. That's a long-winded way of saying it, but yes. Yeah, I mean, I often think in terms of love your neighbor as yourself.

And I find that that same kind of logic works in systems. so if you uh if you just treat somebody as uh as a vendor or uh just say okay this is just the service you're going to give me this is how it is then that that's the end of it and you're just using that service as is but if you sort of get into uh into their perspective you can unlock additional potential and it allows for creativity and hard work to flourish in a way that it wouldn't otherwise. So just to kind of give you an example. So some of the generators that we would use, they normally provide 480 volt power. The Bitcoin miners, I'll spare you the details, but in most of these bills, you need 415 volt power delivered to the bill. You can either buy

a big transformer, which has additional capital expenditure to transform that. Or if you know a thing or two about how the generator works, you can get around that. And so what you can do is you could reduce the stimulating voltage on the generator, and then it will output 415 volts with a certain D-rate associated with the generator. So by paying attention to how that system works, that's not necessarily a system you'd think about, you can, you know, save a bunch of money. And that same kind of concept trickles up and down the, the chain here. So when you think about, okay, a miner is plugging into a given set of infrastructure. Okay. Well, how does that infrastructure actually work? How can we, you know, you know, fully integrate in a way that

takes into account all the little nuances. And this underclocking journey that Reid and I went on was born out of a problem that we had. So we had a bunch of pre-ordered S19J Pros that were delivered, and we're trying to get these things plugged in just as soon as possible. Well, we have this legacy site out of Washington State. I mean, this thing is from back in 2013. It's a really old site it was originally designed for s19s way back in the day oh yeah sorry sorry thank you and um and yeah so you have tons of rack space but the pdus are relatively small for you know what you would uh you know today consider a standard mine so uh basically we were uh when we got these S19s and we're putting them on the rack, the stock power draw on an S19 is about

3000 Watts. And these PDUs were rated up to 5,760 Watts. Two of them on a PDU. Yeah. Right. And so we, we start with just plugging one per PDU. It just seems like ridiculous, right? Cause here's this PDU design, it's got eight plugs on it, you know, designed for, you know, for a bunch of S9s and we can only fit one S19. It's like, this is ridiculous. And so then we sort of got clever and realized, okay, as long as all the phases match up, we can put one and a half miners, you know, per PDU. And so you'd split, you know, cause the S19s have two plugs. So you could split one over two PDUs if it was on the same phase. And so then we got, you know, Oh, 50% improvement, you know, but we still had a bunch of miners that were trying to get a home for it's like okay we have

to fix this and i was and i um i was like okay what if we could underclock the miners a little bit and then we could fit two miners a pdu that would be much more efficient and so we we sort of started thinking in that way and we were doing some some testing and it's like hey you know we can do this and then uh you know i i'm between the two of us uh between reet and i i'm probably more of the cowboy and so i just got to thinking well hey wait a minute what if we underclock them even more what what if what if we just stuffed three or even four on a pdu and everyone on the team just like isaac you're freaking nuts like what are you talking about you know we could barely get one and a half and you want to put four minors on a p i'm like yeah i think we can try

this and and this comes from uh you know again having an intuitive understanding of the way this stuff works is really, really helpful. And a lot of that comes from pain as we pointed out. So brains had put out this really great piece of research on, on how temperature can affect the, the watt draw on, on an S19 model. And I was like, huh, okay. And at a fundamental level, that makes sense. So just to give you a kind of a first principle approach, you know, how I think about this. I think of chips as these little cities and electrons are like the little cars driving around inside the city. You get to change volts and frequency. So frequency is how fast the stoplights turn red and green. Those are the logic gates. And the voltage is the kind of

engine you're putting on the electron. I know there's lots of other models. You can think about it as you know water pressure and you know how big your hose is lots of ways of thinking about this is the model that i i sort of use and it's like okay these chips are are smaller so now you have to start thinking about the quantum physics of this because you're at a scale now where uh electrons can quantum tunnel through the logic gates like you're getting to a point where you need to think about like what's happening you know at the electron level and it's like okay well what is temperature to an electron so think about driving down the road in your car and somebody's shaking the road like like an earthquake back and forth and back and forth that that's how an

electron perceives heat because temperature is uh it's vibration at the at an atomic scale so an electron is just trying to go along its merry way down you know the uh you know the path on the circuit board and it gets hot and all of a sudden the road's shaking all over the place and it's veering and hitting on the gas and coming off the gas you know it's very wasteful but if you can stabilize the road make it you know nice and cool on the chip level the electron can get to where it needs to go without using as much energy and so okay uh you know brains puts out this piece of research and says, hey, these chips are very susceptible to temperature. It makes sense. They're a smaller chip. They're more susceptible to vibrations. You know, it's the

difference between, okay, I'm going to be on an eight lane highway that's shaking back and forth versus, you know, I'm going down, you know, a little cobblestone street in Rome, right? You know, like I'm trying to squeeze through because it's a tinier chip. You're much more susceptible to vibration so how can we reduce heat well i mean you could you know immersion cool these things you can uh you know feed it liquid nitrogen there's all sorts of things you can do to keep it cold but one way that people don't often think of is if i run less power through this chip there's going to be less heat right i'm reducing the the heat density because i'm reducing the power density so if you use less power it's more stable from the electrons perspective which means

it should get more efficient and so um you know brains was doing this with like you know stock clocks and overclocking it's like well what if that concept holds downwards as well so what if we could make it more efficient yeah so that's just you know a first principles you know what And we tried and it seemed to work. And we ended up getting four miners of PDU at this legacy site because, you know, we yeah, we were able to reduce the power going through each individual machine makes makes a more stable environment on a per chip level and you get more efficiency. So if you think about your site as, okay, I have X power coming to this site and then Y output, your hash rate, you just need to stuff it with more machines. But then you're getting more incremental hash rate for the given power draw, which means your margins improve. So it's better to think about it in terms of, hey, how can I get more money out of a respective site than how can I get the most hash rate? You know during a bull run. I mean if Bitcoin goes to infinity you can afford to mind at any price

So you just want every hash you can get in a bear market. You want the most efficient built So you want to overclock in a bull run under clock and excuse me overclock, you know Bear market yeah, so now that opens up this whole possibility. What if you can have these flexible clocks that respond to network difficulty and hash price uh maybe you know your ambient temperature so you can start to get pretty sophisticated and thinking about how all these things are affecting your your net profitability um but anyway maybe i've said too much but but go ahead someone else talk i don't think you said too much this is fascinating reid do you have anything to add to that it comes out of necessity right i mean for me you know the the the four

miners per pdu at a legacy site i was like oh come on man like seriously like how do we do that um i'm a visual person and that's where all these maps came out and said okay well let let's look at what this machine does systematically right and and is there a sweet spot is there some way to quantify this it's easy to kind of pick and choose things but if you don't have like a good roadmap to it the data doesn't really mean anything okay glad you got to underclock it's more efficient Cool. Is it the best way to operate? And the answer is not really. So if you can find this channel, so we found that there was a point if you underclock too far, it wasn't as efficient. There was a kind of a de minimis point of saying, here's kind of this area that works better.

And that's kind of the range that we want to operate it. i think that's true on the high side as well we're still i'm still going through step by step on the um the overclocking side it's a little bit different but not that much same concept is that there is kind of this this channel where it works the best and that's you know using the kind of cars and lights analogy is that you don't want to have too fast of a car because then the lights are going to be out of sync. You kind of want to have them in sync and that's going to be the best place

regardless of whatever volt frequency combination you have in the most efficient band. Yeah, so just kind of leapfrogging off of that. To Isaac's point, that the just mapping things out is only the first layer you can continue that and saying what machine mix do you have and you just keep on layering machine over machine over machine and then you start to get the shifting map and saying this is how i want to do it and if you have two different power mixes for example site one and site two you can start optimizing how you

want to place your machines and at what watt per terahash so it could see a lot more you know that that's essentially what we're working on uh is how can i use this fourth dimension of the machine's capability and how can i make a good decision off of it that's essentially where it's gonna go yeah so yeah we're at like the like even though asics have come very far and are extremely advanced now producing a lot of hashes getting more and more efficient on the management side it seems like we're still in the stone age or the bronze age where we're figuring out like it seems like what you guys are describing paints the potential for

like a crazy integrated system that like in the future there will be processes and software that basically like isaac like you mentioned look at difficulty and hash price at any given point in time and then manipulate the operation to be as efficient as possible from a profitability perspective yeah yeah and that that sounds wonderful except then you need to figure out how to calculate uptime because it starts because you've got your hat like oh well our target hash rate was this at this time and something else another time it's like okay you know and most people they're you know saying okay this was our uptime i expect the machine to produce this much hash rate i expect my power bill to be exactly this every month and it's like okay now we're

gonna just you know mess up uh yeah the account well at that point you just change the kpis right like maybe optimal hash rate production per machine isn't the right kpi maybe it's profitability per machine and if you've you sort of target profitability over um optimal or not optimal profitability would be optimal hash rate but uh as much hash rate as possible maybe that's something the industry should should move toward well you know and you have to kind of consider macro conditions in twos like depends on when you bought your machines right like did you buy at the top of the bull market at a hundred dollars a t or did you like do you have cash right now and are you buying machines at 15, $20 a T. Right. You know? Well, and, and, and here's one of the implications

of, of, of this, you know, from a, you know, minor allocation standpoint, like, you know, what do you want to buy? I mean, you know, today, uh, like an XP goes for about $6,000 and S19J Pro is going for about 1600, 1700, right? So what if, uh, what if you can underclock an s19j Pro and get not quite as good but get comparable efficiency to an XP very similar all right do you want to spend six thousand to get that level of efficiency or do you want to spend you know four thousand dollars to get something of comparable efficiency and comparable hash rate and now the downside is you need to rack spaces instead of one rack space but presumably

you could uh you know get that rack space for less than what the difference is you know which would be two thousand dollars so in terms of hey what machine are we going to buy uh you know it you know if you use techniques like this it can unlock um you know capital allocation in a way that you may not have looked at so yeah it's just about saying you know intuitively what is each system capable of and how can they integrate well uh because you know someone who isn't using this kind of you know firmware take and they just want the best efficiency they're going to buy the xps but if you can you know kind of use this firmware hack uh you know then you can get that for for less right now that's what it's all about doing more with less there's going to be whole teams

dedicated to figuring out like the the correct capital allocation rack space yeah because there's so many there's so many variables yeah correct is electricity cost hash terahash per machine it's it's insane to think how complex is it complex yeah it's complex it's a complex system like how complex this is going to get and how how much brain power is going to be dedicated like you're gonna have a whole like trading desk trying to figure this stuff out you know or or chat gpt you know we'll see just ask chat gpt yeah make sure you ask it in the right way i mean i wouldn't be surprised within a few years one of these ais is a you're able to be like hey here's one here's what i'm trying to fine-tune for yep yep mine gpt trademark you know but that's

kind of the the thing of where you like this is the difficult part right is that do i buy more rack space do i buy more power capacity do i buy more machines like it's all a process and it all has to kind of work in this this um flywheel um sequence right like you can't have more of one thing and then not have the other otherwise you get you know kind of behind the eight ball and you want to make sure that you know you might go make deals that you don't like or or something else but if you kind of use this as kind of a tempering mechanism saying what is rack space really worth right now containers rack space it's cheap because nobody's buying them but the long term value of them i personally i think it's severely undervalued right now rack space is

severely undervalued yeah and what you're getting at is that like these decisions are dynamic they they're changing with every block that's produced and with every every tick of the the price moving up or down and so you have to have the ability to make the correct decision in a dynamic way on the go depending on the market conditions yep and and i think eating smaller steps so instead of saying i gotta go build a hundred megawatt facility it might be hey i'm gonna make these decisions in smaller increments and have the ability to do so and then kind of roll it in smaller chunks,

which actually, if you take a 10,000-foot view, de-risks a lot of those decisions, I think. Yeah, that's one thing I've become partial towards over the last two years is I think these megamines are not going to be the wisest way to build out a mining operation moving forward. I think taking off smaller 5, 10, 20 megawatt chunks at a time makes much more sense.

Yeah, I mean, no problem with trying to go get 100 megawatts worth of power capacity, but like the infrastructure and what goes into it and like when you buy machines, how you buy machines, and then you start filling into it, I think that process should be, I don't want to say compressed, but kind of like concurrent. it's really hard to say I'm going to buy 20,000, 30,000 S19s a year and a half ago you have to develop a just in time

process as soon as the electricity is up the rack space is up and you're able to plug in you should have that project management set up where as soon as we electrify we are plugging miners in

yeah I think that the the underclocking play is gonna have I think it'll have outsize effects on profitability relative to a more distributed model and what I mean by that is if you look at a mine that's on grid you know one of these mega mines right they sort of have no penalty necessarily I mean every sites different, right? Sometimes you have a minimum offtake, what have you. But if you say, all right, well, I'm allowed to reduce how much power I pull from the grid without penalty, then in a bear market, when hash price drops below whatever point the stock firmware machine would go unprofitable, well, what if, hey, I can underclock now, increase the efficiency on a per

machine basis. I'm reducing my overall power use, but I'm still profitable as a site. That doesn't quite work the same if you're off grid, because in that situation, if you're renting a generator or you've purchased a generator, you already have that as a payment, right? And that doesn't change. It's sort of like buying a generator. You can think of it as having committed to a minimum offtake agreement. I'm going to buy exactly this much power. And it's pay or take, if you will, right? So you're committed to using whatever the capacity of the generator is. So in that context, sure, you can make the machines more efficient. But if you're using less power, you're just under utilizing something you've already paid for. So it may not make sense unless you can somehow

pull generators off, or maybe you could bring another container on site and use more of the generator, right? So it's more sophisticated. You can still use it. It's just, it requires more playing with the physical infrastructure. And on grid site, they can push an underclock and they just stop using as much power from the grid and their margins are improved on a per watt basis. So a little bit of nuance there. Yeah, and now we're dovetailing into Bitcoin mining, changing the game of these PPAs, how they're structured, what is allowed, going and educating utility side of the equation. Like, hey, we're going to do this weird thing that probably no other customer you have wants to do, but trust us, it's worth it in the long run for you. Yeah. I mean, in terms of integrating with utilities, I mean, that's a whole other can of worms with the firmware side of things. I mean, so right now, the way it works is if you want to participate in demand response, you're just turning off machines. I mean, that's how it works today. You suspend mining, or if you have a smart PDU, you just turn the miner off altogether.

but you're you're you're basically keeping your same rate of efficiency uh when you participate so so if you're you know let's say okay you know you're um you know whoever manages the grid whether that's an rto or municipality whatever says hey we need you to turn off half of your uh your power use is half of the miners get shut off and the other miners are still running at the same efficiency but what if instead you said okay i'm going to reduce my power use by half on a per machine basis so during that curtailment you just make all of your machines more efficient so so you're increasing your margins in that curtailment that that's more money juicing your margins right because you're still mining with every miner more profitably and then

you're getting paid correct turning down yeah right so so that i i think that's sort of where we're headed is uh having dynamic tunability on a per miner basis i i think that's gonna to create huge margin boosts for a lot of these on-grid sites. That's going to be massive. Which is not even demand response, right? Well, it is a demand response, but it's not like a binary decision at that point. Right. It's a dynamic response versus a binary response.

You're scaling your hash rate up or down, and so you're responding to the request to curtail, but you're doing on a per machine basis and as you do that on a per machine basis you're increasing your efficiency so you get more hash rate out of the power you are allowed to consume at that time so yeah i that's where we're headed i think this is like some sci-fi shit here yeah and half the time you know reet and i have conversations like this like hey can we can we do this kind of thing uh i mean one of the things that uh we've been talking about recently is hey what if we could be designated as like a virtual power plant so

we're a buyer on the per kilowatt hour market we're a seller on the um on the capacity market and so you just uh you know if the grid is asking for more power you just turn off yep you know you don't even have a power plant on site it's just direct tied to the grid and you're anyway we have these crazy conversations and you know nine out of every ten just die on the drawing board because It's just too nuts. But every now and then something comes through. You know, I'm just a simple miner. Let me mine.

I think that's what we're getting at. Like, yeah, mining is much more complex. And especially if you want to be profitable and in it for the long run, like you have to have these types of discussions and this train of thought. And I think, you know, I'm really thankful for Reid. I think we make this really interesting team where I'm just sort of this cowboy, like, Hey, let's just throw stuff at the wall and see if it sticks. And, uh, and you know, he, he tolerates me enough to hear me out and then we'll calmly explain why I'm

just so wrong. And it, but every now and then he goes, well, wait a minute, maybe there's something there. And then, yeah. So, so like, Hey, can we look here? And then he just dives down on the, you know, like an electron microscope microscope and says, Hey, did you think about this? Like, Oh no, that nukes the whole idea. Nevermind. and then we just continue on but it's just sort of uh yeah this great interplay i'm just pushing the envelope and he's like whoa whoa whoa and uh yeah sometimes we fight like an old married couple about stuff but we always yeah it's great that's an understatement but yes yes

ideas have to be able to be executed so when physics are defied then i will put my foot down but um you know when you get into this i think there's so much potential in this uh firmware space and unlocking the potential of of the machines i mean that's just one step right like but you still have to make the uh environment conducive that's where bitcoin comes into a physical it in contact with the physical world so if you don't make the the mining machines happy they ain't gonna work for you so you know we can do all this this fun stuff but like as part of uh what we wrote like i could have the same volt frequency uh you know uh on the graph so to speak um but if that temperature is way up here

doesn't matter right and that's that's controlling your environment that's making sure that everything kind of works in concert in the way that you you've hopefully designed the system yeah that gets to another point like with all these efficiency gains and the games that you play with voltage and stuff like that it's going to be hard to tell because enough time has not passed yet but do you guys have inclinations on what this will do for the the life cycle of these asics do you underclock them it's i don't have the back data but my gut instinct is uh it will you know expand the life of that machine as long as you don't have you know dust going in and and mucking it up like right yeah i mean it's it's a very

simple concept, you know, take care of your things and they last a long time. It's very simple to take care of it and then figure out what it needs to run well. I mean, just as a, you know, first principle concept, I look at the, you know, the S9, you know, when brains came out with this great firmware, it was an underclock, right? But that extended the whole life cycle of an S9. And I think i think what we'll see is that trend will continue with with lots of other miners where underclocking pushes it into a more efficient regime and then you can afford to run it longer um so yeah i i yeah i think it'll change yeah it'll change a lot yeah it's crazy to think that the average life cycle of an asic before the s9 was like 18 months now people are still running

s9s today seven eight years eight years almost later yeah i mean but now they've modded it to a double shoebox like one power supply for two like yeah i mean yeah you know looking at um looking at what bitmain did with their s19 i really respect sort of their take because you look at the thing it's like why does this have two plugs and then if you think about it it's like well they've got their own mines as well and all of that infrastructure and including their customers too right it's all designed for s9s and you say okay i'm going to double the power use on a per machine basis am i going to put put a bigger plug am i going to put that c19 plug and make everybody gut their mind and and rebuild all of their

electrical infrastructure or am i going to put two plugs and anytime you pull out one s9 or sorry You pull out two S9s, you can put in one S19. It fills the same space on the rack. It works electrically, right? And so thinking – I think manufacturers are starting to think more and more about the infrastructure side of this. You know, where are they going to – where are these things going to go? And, again, that gets back to the concept of loving your neighbor as yourself, right? Like, you know, where is this going and how can it be used well?

um i mean bitmain coming out with these new three-phase power supply uh sorry uh yeah ps i can't talk pdus um yeah i think that's going to be game changing for a lot of folks right because when you build out this this mine uh you don't necessarily have to do the same amount of work you know pulling all the phases apart you can just deliver the three phase right to the pdu and then it gets you know the splitting up gets taken care of at that level um you know stuff like that where it's thinking about how it all works together it's going to become more and more important and to that point just due to like the limits that the asic industry is beginning to brush up against do you think on that side the psu pdu side we've reached like a point

of commodification where you don't have drastic changes from one model to the next it's everything is size really i mean you can make these really big big things more amps more you know it's going to be really interesting and where how the power side is going to to occur for these machines whether or not they're going to be like hey we need bigger and bigger psu's on these things if that's the case we're kind of at the the standardized plug limits already um can't do it so one thing that you can do is go to three-phase power um you will need to modify your your farm a little bit for that but i mean i can still use 14 gauge power, sorry, 14 gauge

wire and push 8,000 watts through it in three phase. So... Sort of reaching that point, is that what you're saying? Yeah. Yeah, I mean, like, what are you going to do? Put a dryer plug on a miner?

I mean, where would you put it? You know, there's just, we're reaching sort of the limits of what electrical engineers had thought up for quite a long time. I mean, Bitcoin mining is pushing the yeah with these um you know the specialty in industry right um like like they're just there is not a computer plug i mean maybe the l7 but like in general you know the c19

is your limit it's 20 amps right i mean if you want to go bigger than that uh i mean if you want to go higher voltage than that then you start reaching problems with you know arcing uh which is why everything is rated 250 volts or less right that's a you know an electrical standard so um when you start to say hey i want to increase the voltage for efficiency or hey i want to put more more amps through you're reaching the limits of uh you know what's permissible under electric code like there just isn't a plug that can handle these things so you have one of two options either you put you know as these chips get more and more power hungry or you know they're putting more chips on a board as I get more efficient. Um, you know, if you want to use more power on a per

machine basis, you've got to go three phase, you've got to start doing direct wiring right into the unit. There is no plug, you know, stuff like that. Yeah. The other way to handle it is you just put less chips per machine and you keep the power, um, you know, the form, the form factor the same, right. Because of the limitations in electrical code and like what's physically possible so i i think you'll start to see um yeah i i think either you'll start to see more three-phase stuff come to market or you'll see people using the same sort of power form factor and they just put less chips yeah god you two are so fucking smart thank you for the education you've given me uh over the years that we've worked together i think you guys the two of you

austin tom have really helped level up my knowledge of the space it's crazy but we've you know and that kind of boils down to the fundamental thing like the team matters right if you have a lot of people or good people to to kind of collaborate with and educate and kind of lean on. You're golden. Yeah, well, it's just, you know, it's, it's, it's staying humble and working hard. And, you know, part of being humble is listening. Just recognize that, hey, you never know where the best idea is going to come from. And just being open to it. No, I think that's something you've done really well, Marty. I think that's why you're such a captivating uh you know uh podcast hosts you know it's like you just you you you know

where uh yeah you know where to let people just run when uh when they want to go all crazy so i don't know well i don't think it's crazy at all thank you for the hosting compliment but like no i i'm being earnest here like talking to you guys is multi-hour long phone conversations trying to understand what the hell you two are working on significantly up to my game personally i mean i i assure you like we feel that way most of the time so it's not you know like there's so much of this stuff that is like you know a master electrician you know they'll spend their whole life devoted toward understanding all these crazy nuances uh you know, harmonics and, you know, you know, phase shifts and, you know, there's all this stuff,

right? It's a very, you know, sort of enshrined discipline. And it's like, you know, networking and data management, you know, you know, you know, network security, HVAC systems, you know, structural engineering and welding. Like there's all of these crazy disciplines that Bitcoin and Bitcoin mining pulls on right down to economics and, you know, political theory. I mean, there's all of this stuff that just gets sucked into it and so you're never bored but um but often overwhelmed and feel like the stupidest person in the room but it's just a matter of you know i i often describe bitcoin mining as just uh you know just enduring pain you're just like hey i'm gonna just uh yeah that's why i say you gotta stay humble if you got yeah yeah very masochistic

well it's not it's it's always listening i mean you say stay humble stacks ads the stay humble is is listening to your environment listening to the uh listening to the market listening to what is needed to make this ecosystem thrive and so like i said i'm not an expert i am i just know what has caused me pain right and and and that just keeps on stacking and stacking and stacking so i can say hey i know that you want to go this direction but really what you want to do is go this way yeah don't don't make the same mistakes that we did or hey by the way there's a pitfall right in here it seems like a good idea but it's not and that's just experience so the more people we can get in here to get that kind of experience into this like

i'm an open book i i'm happy to bring more people in that's that's what's going to accelerate adoption and just make this ecosystem work would it be i mean would it go too far to say that like bitcoin mining as an industry considering all the different variables and core competencies it pulls on could incite like an engineering renaissance wow that okay that is um good well i because what what bitcoin mining is doing currently is it's disrupting the idea of a specialist because if you just have a an electrical engineer specialist and an hvac specialist and a you know data networking specialist i mean sure you could go get all those specialists and get

them in a room together and then like it's going to take a while right to get everybody on the same page and what you want to go build together because there's a lot involved in these in these sites or you can have somebody that has basic competency in each of those respective areas and they sort of they're they're more like a conductor than a musician right they allow each of the specialists to do what they're really good at and and they they have an understanding of the other disciplines, right? And so I think Bitcoin mining is bringing back the Renaissance, man, you have to know a little bit about a bunch of stuff, if it's going to work well. And it's great to have specialists in your court. But a specialist, you know, on their own is not going to

be able to make a good mine. I think the specialist is absolutely there. But it's easy, especially is say if you're a civil engineer you're going to talk to some other civil engineers you're not really going to talk to the the other guys and if you do it's through a consultant or through some manager project manager i think having a diverse skill set team is very important especially if you have a lot of different experts in the uh from those fields but you need to make them to work together towards a common goal and if everybody well i mean hell if you believe in bitcoin and everybody believes like that's the future then you already have a common goal in it's just about working together at that point so if you can kind of unite around the

progress um platform absolutely i think you can have a lot of different disciplines working together and and as they do already but more integrated in that way yeah fuck i'm bullish it's uh always has happened we've never lost that bullishness no no but it's again like sitting down and putting this all together like the random bitcoin mining could be bring back the renaissance man in the engineering sector i think you have to like you said i really like the conductor analogy like you have to be able to understand how the different parts interact with each other yeah yeah i mean if you're always playing the same song you know every week week in and week out you know uh

eventually you don't need the conductor right because the musician just sits there and they're doing exactly the same thing off the same piece of sheet music it's like it doesn't change so you go okay well we don't need the conductor but once you start changing the song and like you you uproot the the status quo on the kinds of engineering you're doing because it's very dynamic there's lots of systems at play here and so what you know once you start doing that you need the conductor back to keep everybody in sync otherwise it's just a mess uh i think i think this i think this rip is going to be very popular amongst the mining industry because again i think a lot of people like have a rough idea about all these problems you guys are describing

and as you said you're not experts but i think the experience that you two have goes far beyond a lot of the individuals in the space i think that i i mean we saw that too when you guys launched the thread on um the research that you did on the underclocking the response to that was like holy crap i can't believe somebody did this thank you for sharing it i guess that's another thing to touch on like why share this isn't this alpha um shouldn't you hold this close to the chest that was some of the feedback from the thread you know we thought about it in that way but it i think it's hard for if you're trying to promote this promethean

existence like you want to share the platform to do it i'm not telling you how i did like how i'm writing my own algorithms or all that, you need to go put in the work. Sometimes it's just about showing, hey, here's the possibility, right? I'm not perfect. I might have blind spots, but I would respect the guy who uses that as the platform to say, hey, I looked at it in a different way. how about this and if he shares that too then it it perpetuates that that cycle of sharing mutual mutual progress i mean that that's how a renaissance happens you need to get people talking and if everyone's you know covering their butt with proprietary jargon and and patents i mean the the industry won't move forward right um i mean like i remember uh you know when i first

got into mining i mean there were no youtube videos there were no tutorials on on how to build these things it was just like okay i gotta just google a bunch of random stuff that sort of has proximity to what i'm trying to do and i mean that i remember back when i was building gpu rigs like there was just no information on this stuff it was just a lot like just sheer pain banging your head against the wall day in, day out for months before I figured it out. And there were a lot of people in the space who had built before and they were showing people how to build these things. I mean, go right back to Satoshi himself. He made it open source. He put it out into the public domain and said, here guys, that's what I'm trying to do. This is what I'm trying to build. How can

you help me make it better? And that's really the spirit of a Bitcoiner is we're going to build a better world together and to do that you know you can't always be worried about oh someone might be trying to stab me in the back like you know but part of this is hey we're all in this together you know as as bitcoiners and we're gonna optimize and build better together so yeah um and it's just about staying at the tip of the spear it's about saying look we're gonna be thinking about these hard things thinking about these problems and and try and be among the first to adopt these novel techniques and and technologies and if you just are constantly staying there everybody improves while you do well and uh you know that that's a better world that's how you build a better tomorrow

hell yeah we're gonna win always what uh together we win this is true what uh what should we send the freaks off with any final thoughts anything you guys want to touch on before we wrap up here just test just just work on stuff if something interests you it's possible it's you might not know how but that's part of the process so you know there there are a lot of resources a lot of parallels that you can draw i mean i i got my hot tub heating loop from a wood boy boiler stove idea, right? Like, okay, okay, that's the way I should design it and go from there. So there are a lot of things that you can borrow from. Just share. You know, I like the idea of sharing on some of those ideas and just go test it. I want other people to have this graph and

make their own graphs so that it's like, oh, well, they did it a different way. That knowledge sharing i think is is is what i want to see out of that yeah i i would say um all hard work yields a profit and sometimes it takes a while but there's um there's something that comes from excellence in in any discipline if you do something well you will um you will inevitably discover and it's just about being curious and saying hey you know what are the things that interest you and saying even if it's 15 minutes a day i'm going to just walk in this direction because i'm fascinated and it's just you know i i remember my first gpu rig uh you know i i tried and tried for a couple months i couldn't quite get it you know i was teaching myself how to build a computer

here and uh i i remember i stopped for a couple weeks and i just remember thinking no i decided to do this i decided i was going to figure it out and i went back to it and i finally got the thing going and like there's nothing so sweet as wow like this was so hard and i got it you know i it's like you know this you know i i think about who's the first guy that made fire and like how crazy that must've been. Like I did it. I figured out how to do this. Right. Think about how long that must've taken. Like, like literally there's no YouTube tutorials. There's no one, like somebody had to go figure that out for the very first time, you know, and like that, that's hard. And, and each one of us has this, this sort of personal journey. You know, I've got, I've got

two young boys and there's just this like crazy light in their eyes when they figure out how to do something for the first time you know like oh i figured out how to turn the light switch on on my own oh i figured out how to unlock the front door oh shoot mommy uh please you know like that but still there's this there's this beauty um yeah in that and so it's just work hard and you know stay curious and don't um you know the future belongs to those who believe in their dreams so just don't give up and marty thank you for what you do and and giving people like us a voice right and being curious about what what's going on because you above anybody else understand that it takes so many different people from so many different disciplines to make this all go

forward pretty much yeah i mean i was beautifully put by both of you and the thanks goes both ways it's been a net benefit massive net benefit to my life working with you guys and learning about all this stuff and I'm extremely happy I can't believe it took this long to get you guys on the show I know I'm extremely happy

it has finally happened because I think there is going to be a lot of aha moments for people in the mining industry listening to this and like you said again Bitcoiners giving back really trying to live close to the ethos that Satoshi

he set forth when he launched the protocol is what's going to create this cascading effect. I do think particularly in the mining industry, there is a bit of keep things close to the chest. That's your alpha. But I think considering the enormity of the problems that we're trying to solve, I think this will have a positive feedback loop sharing this information and

really putting this out there. So thank you gentlemen for doing it. Pleasure. We're in it for the plebs. hell yeah well it's been fascinating we'll do this again I'm sure

it's not going to take five years this time I hope not oh my gosh has it been that long I'm going to be all grey by then if that ever happens you'll be all grey and I'll be joining you in the bald crew hey bald is beautiful

it is don't let anybody tell you otherwise all right gentlemen thank you thank you gents cheers seriously thank you you guys are fucking crushing it hey love you marty love you guys too peace and love freaks it was fun isaac you want to get it oh i got this here now peace love freaks Thank you.