Squeezing Profit from the Margins: Uncovering Stranded Sats in Mining Operations | Bitcoin 2026

All right, good afternoon everybody. Happy to see some faces here. I hope you uh been enjoying the conference so far. Uh it's been a lot of side events. So the third day is always a bit slow, but I'm happy that you uh joined the session. Uh my name is Nikico Smith. I'm the founder of Digital Mining Solutions. uh and we help capital allocators uh investing in uh digital infrastructure and we also support uh operators in scaling their business from 5 10 megawatts to a full-fledged infrastructure platform. Uh and today we're going to talk about um operational efficiency that the title is very fancy but it's all about efficiency today. Uh and we're going to talk about efficiency because uh the margins have been very very tight in 2026. uh we've been below the $40 per pahash per day uh looking at hash price and that's uh it's tough. Uh so in this tough environment it's uh yeah the efficiency operational efficiency is not a it's not an option it's a must and uh many miners have been upgrading their fleets um to to the yeah with the with the latest generation A6 let's say uh some miners have been uh moving their equipment out to other places where power is cheaper. uh the bigger miners you see all pivoting to AI HPC to get more out of their power and infrastructure. Um but we believe that's there's much more to do especially when you look inwards to the operation that you already have. Uh there's a lot of ways to uh to optimize your operations uh by using firmware the right uh management software uh ensuring that uh you minimize up time downtime. Um and so yeah today we're joined by a by a group of experts. Uh welcome guys. I wanted to do a short round of introductions before we go into uh to the content. I want to start with you Dan. Uh you're the CEO of ASIC plug and and god farm god. Uh and with farm god you you focus on real time monitoring stability uh optimization operational intelligence and I was wondering uh what do you think is the single biggest operational mistakes you see uh miners making that are eating into their margins? >> Yeah, absolutely. So the the biggest mistake that we find out there in deployment is having the wrong tech stack and not having the technology aid the on-site personnel in the most efficient way possible and that results in additional personnel needed um more manh hours than required for the operation and that's usually your biggest cost. Power and labor is your highest cost. >> Well, it's good to to hear that it's not AI taking over. We need people still. Um, next up is uh is Bradley Peak. Hi. You're the head of global sales at Venice. >> That's correct. >> I think uh most of us know Venice. You're a legacy firmware provider. Um, and I was wondering Bradley uh what exactly does uh custom firmware uh do that um that the stock settings cannot? >> Yeah. So, when you buy a bit main amp miner out the box, it comes with stock firmware, right? So and then you have the option you can either keep the warranty or after a few months switch to custom firmware. Venicia is the leading one at that. Um and it really solves the problem because stock firmware is a one size all fit solution. So it needs to account for a wide variety of say chip variance, climate differences, power quality across across regions and ven um its main product is obviously perch chip autotuning. So this means that say a client has the target of the highest hash rate or it could be a minimum watt per terahash. Uh it could be the best balance for site and power. Um veniche allows that granular detail and attention which is why you can get up to 20 25% efficiency gains when you're using custom custom firmware like ours. Yeah. >> All right. Thank you very much. And and last we have Alexander. Alexander is the founder and CEO of Interhash, a company that specialize in professional crypto mining that provides scalable and sustainable mining infrastructure. Alexander, uh, from a perspective of a mining operator who, uh, who is m is managing uh, machines at scale, um, where do you think you find, yeah, the the most consistently overlooked source of lost revenues? >> Hello, thank you. for my opinion there's a lost uh revenues it's everywhere and mostly um it's um lost it in inefficient way. So a lot of people when they building their farms they lose the small parts how to organize all the scales all the networks uh in the data center also it's about employees also it's about uh software you use pools firmwares so everywhere it's a small small small pieces of u your lost earnings and if you will get a discipline if you will work hard from each part. So with a with a very small pieces of the task in this case I think the efficient of the farm can increase by I don't know from five to 20% for my opinion or maybe even more because uh mostly people looking only on the electricity cost on the prices of the kilowatt but uh it's not the main uh reason to be successful also we can check that a lot of miners lose their money in the maintenance. So mining business it's you know like it's a lot of uh thing that you have to do and there is not uh there is no uh not important things over there. Well, that sounds promising that there's uh opportunity everywhere which also is challenging obviously but uh uh Bradley uh I want to move the discussion to firmware and hash rate optimization and I think it's obvious that we start with you uh could you tell us a bit more like when you apply uh Venice firmware what is happening under the hood and and what kind of results uh are you getting when you apply firmware? >> Sure. So I'm not sure if everyone in the audience has heard something called the silicon lottery. Um but essentially when a chip is produced um it varies in quality. So what the manufacturer usually do is put it on a board test it and see whether it fits within their desired range. Um if it's too far away from this desired range they do something called chip binning. Uh so they'll use those chips for cheaper machines essentially. So because of this silicon lottery um stock firmware doesn't really quite hit the mark uh because it's sort of offering this umbrella for all of these different chips because um Ven offers something you know per chip autotuning. It means we can create profiles for every individual machine and then we can say overclock stronger chips or underclock those that need greater stability because they're weaker, right? Even though they're the same machine. So these are the sort of granular details that we can control at finish leading to better efficiency across the board. These um profiles take anywhere between 3 to 4 hours to create but once that's done they automate this process. So they'll change they'll adapt pretty much automatically. Now Venish as well it's not just firmware products we've got an entire ecosystem of products that not a lot of people know about. So of course we're the largest firmware provider. We also have the hashcore toolkit. So this is our farm management product. Helps with things like device setup, diagnosis, uh bulk configuration, things like that. Then we've got pulse. This is our monitoring product as well. So this is cloud-based and gives greater visibility. And then we've also got the hash core proxies. This a stratum proxy that sits between the pool and uh our firmware. This really key um because one single pool can be quite unreliable, right? Um so if that pool goes offline, you need to have a backup. This is called failover which is something we support at vanish niche 2. Um and also we've got lots of partners. So if you have heard of brains probably you know that they'll probably ask you to use their pool right. Um we actually have over five authorized partners uh pool partners right now which means you can actually get pool fees scraping zero in some instances. So we've got a wide variety of partners. Um so using the whole Venicia ecosystem is definitely worthwhile. And all of this together including our firmware gives us much higher efficiency and the best efficiency in the market when compared to our competitors. >> Thank you. I for one wasn't aware that it was such a broad ecosystem. So it's good that you shared that. Thank you very much. >> Uh Dan, I want to go to you. Underclocking and uh low power mode has been um an option for many miners to survive these periods of low uh low margins. How do you look at that tradeoff between uh um yeah maximizing for hash rate and optimizing for power efficiency and to what extent does the the rate of your electricity influence a decision? >> Yeah, great question. So the electricity rate is probably going to be the number one driving factor of that. However, the other aspect is what does your PPA say as far as your required load? If you have a PPA that requires you to take off a minimum amount of power and when you go to optimize and underclock everything, if you don't hit that, sometimes you're going to have to crossbalance and, you know, overclock maybe your more efficient machines while underclocking your older generation so that you still hit your minimum load consumption. You can curtail back everything else to maintain your PPA, but squeeze the most hash rate out of your kilowatts that you're you're getting in place. you know, we we manage tens of thousands of miners with firmware u just like manish and it's incredibly um efficient to be able to say I need to make this one tweak and then see it go across the entire fleet and dial in exactly how you need it for that day for your demand response, your curtailment or the es and flows of the energy market on the daily. >> Well, thank you for that insight. Uh, Alex, uh, this this dynamic way of mining uh, uh, is is becoming more and more popular. Um, from an operator perspective, how do you plan a buildout when you have this uh, variability in in power draw and cooling requirements? >> So, dynamic mining is very uh important right now, especially on this bare market where we where we are right now. And when we try to build the data centers, we also remember that we have to be flexible and we have to put this flexibility between the high performing and also when we do down walting. So some from from time to times uh we understand that uh the mining h have to be fully efficient also uh when it's not fully allowed. This is why it's helping us. So the planning and the flexibility. >> Okay. Um Bradley going to you. Um you see that many of the latest AS6 that come out they already come with a with the eco mode, normal mode, turbo modes. So I was wondering uh if this becomes an industry standard that there's these different modes already like where does uh vine add value? Uh >> yeah so this this is a really interesting question. And I think the broader question here is uh will stock firmware eventually become so good that there's no added advantage by adding custom firmware. It's a good point, but a lot of people don't realize that we hold the IP for custom firmware. So essentially they can't do it. We have five patents actually since 2021 with another five pending. So if there any of our competitors in the audience then they might be sweating a little bit. Um but it's the truth. Um, and ultimately the current turbo ecosystems, these are just two approved channels and they really don't change much. They enable you to at one click uh overclock your machines, right? Uh, when you've got, let's say, uh, low temperatures or price might be low and do vice versa. Vanish is perchip autotuning, so it's much more granular and much more effective. Um, I think if anyone here is a minor and they're running stock firmware, definitely come and speak to me or one of my team after the uh conference or even after our speech um because there are so many different ways that we can actually improve firmware against um our competitors especially with Brains uh stock firmware. So, it's definitely worth looking into and I think the main point here is no, they can't. We have the IP and we've got 10 years of de developer experience behind us so we really know what we're doing. Okay, thank you for that. Uh, Alexander, uh, Interest has a close relationship or partnership with VIA BTC. I was wondering uh in your view uh do miners put enough emphasis on testing uh the actual pool payout performance in making that choice for pools or is it uh is is it mainly driven by the headline fees? >> Yeah. Uh so for my opinion a lot of miners especially small one right now they're focusing on the pool fee or the brand recognition it's true but uh if you're running large scale data centers and if you're you know pro in the mining you have to look not on the fees but more on your earnings. So this is why in our company we also try to compare the profit per terahash. So how many satoshes we will get for we will get for one terahash also. So to increase performance it can be good to build strong relation with the pool and sometimes uh the pool offering bundles with the firmware it's also works. So in this case you can uh increase your income make lower fees and uh make your mining efficient more uh more successful. Okay. Uh then I want to move over to operational efficiency. Uh and Dan going back to you uh uptime is one of the most uh critical metrics in mining. So I was wondering uh how does uh management software effectively track and optimize uh uptime at the facility level and then what do you think are most common operational failures that uh that impact this uh this uptime? >> Yeah, absolutely. So um our solution farm God tracks all of the data from the miners themselves, aggregates it into a single pane of glass management, allows you to send, um commands, write rules to correct or or circumvent problems that come up as well as track and log into our ticketing system. So when you have a minor that maybe three, four, five times a day it reboots, but it's not often enough that the technician on site is really going to notice it himself. it'll start saying, "Hey, we're noticing a pattern. This is a problem. Let's create a ticket and look at it." So then they see in their their ticketing system some comes up, hey, we've rebooted five times in 24 hours. That shouldn't be happening. These were non-cretailment events. And that can help identify there. Um, as far as the optimization and efficiency side, we have the technology out there that we've deployed to run uh AI analysis on minor logs when there's a cause for concern. So instead of a technician now seeing minor in slot 1 2 3 4 5 is down, I'm going to go take it, remove it from the bench, do a test, see what's going on, it'll say minor at this slot is down. Here's the AI analysis. Instead of starting at step A to B to C to D, it can say start with step E. This is most likely what's going to be your issue based on the log data. And when you have that all done in micros secondsonds automatically rather than your technician having to get into the IP of the device, read the manual logs, scroll through and find out what's happening. It it's a significant time savings on the human capital side. >> All right. Uh Alex, um you know, there's a lot of hash rate or hashboard failures, uh fans that uh that break, uh PSUs that die. uh these repair costs can uh can escalate quite quickly. So I was wondering uh uh how do you go about deciding whether you go into a repair or retire a machine and then maybe also uh uh does this decision uh differ where wherever we are in the market cycle. So now being in a bare market. >> Thank you for this question. So it's true. So when we take a decision about the repair so we try to keep in mind a few things. The first one is the generation of the equipment. So what is the potential to sell it? So if we have got uh good opportunity in future to sell it, we uh we will fix it also. Uh so it depends how much uh time will take the money returns for this repair services. So if it will take for example 6 months so we have to think strongly will we do it or not. But uh on our opinion we can invest more money for the repair on the bullish market and less on the bare. So because right now also if the generation of the asex is not so fresh so better just to work with low hash uh mode over there or just to through it. So this is why it's a very interesting and difficult question and a lot of data centers uh do a lot of mistake uh with the repair but uh the ideal strategy everyone have to choose by their own. >> All right and Brad I can imagine that Vines has a lot of data on on the effects of thermal stress and and uh the component longevity across different tuning uh profiles. Can you show share some of those insights? Like >> so I was waiting for this question because something I get a lot is um does Viche affect the lifetime of my machines? No, it doesn't. If used correctly, okay, if you download it on our site and then overclock all of your chips manually overriding everything, including our own developers advice, then obviously you are going to affect the lifespan. But if used in line with our developer recommendations, if you use the machine as you should do, just like you should do on stock firmware, it actually can outlive the um well lifetime or lifespan of that uh equipment anyway as written in the documents from from Bitmain. So I think um it's important to say that naturally higher performance profiles put stress on a machine, right? Then you've got balance and efficiency profiles put less stress on a machine. Everyone knows this, right? But with autotuning, you can look at weak chips and then tune them down. You can take strong chips, tune them up. You can manage fan responses, overheating as well. And this stuff all increases the lifespan. It reduces uh all of these issues with thermal stress and longevity. Right? So, I'll repeat again when used in accordance with our developer recommendations, the uh machines on Veniche will live the same lifetime as they would on stock bitmain firmware. Okay? And also, you got to think about our market share, right? um in the market. According to Cambridge's most recent digital mining report, we're looking at about uh what is it? Uh 26.4% market share. Um that's when we consider stock firmware as well. So we are the largest firmware provider in the market by far. If you compare us just with say uh custom firmware markets, you're looking at around 85% market share. Now it is impossible since we've been working since 2017 that using Veniche would damage the lifespan of these machines. So it is a complete myth. Uh so that's what I'd say. >> All right. And then maybe in line with this question and maybe a question that you get often as well like uh if you apply custom firmware uh you often lose the warranty on a machine >> the warranty. Uh so I was wondering what what do you advise to miners like uh should they uh apply custom firmware and and just ignore the warranty or or is it better to stick to that or hold on to that one-year warranty and uh and maybe have a a machine operating a bit less efficient? >> Well, you need to make capital decisions here uh economically valuable ones too. So, I can understand why you would want stock firmware and the warranty for the first couple of months because because of the silicon lottery I've spoken about, you have chip variance no matter what. Your your machine could break in the first month regardless of how new it is or efficient it is just because these things go wrong, right? So, I can understand why you'd want the manufacturers's warranty maybe the first few months, but then again, it's still a long time for them to pick it up and drop it off. So, you're actually much better with an aftermarket warranty that covers Veniche, too. So, a better economic decision would be to move your product straight to Veniche and sign a contract uh with someone that's going to then insure it. This sort of aftermarket warranty, get the 20% efficiency buff and earn much more money. In this market, it's a uh it's a real no-brainer. You have to have custom firmware to stay competitive. And warranty is not an argument when you have aftermarket firmware, too. And I've met so many people here that are even offering great rates for aftermarket uh warranties. They pick it up um much quicker than the uh the stock firmware does. So it's great. Um I would say again that with you you're looking at an upwards uh of 24% increase in the hash rate obviously when you when you overclock. But this is a massive difference and our fees are so low. In some cases we're you know offering 1.6% to larger firms. Um, and that comes with next to no pool fees when using one of the authorized providers. So, there's really no reason that you should be using stock firmware. There's absolutely no reason you should be doing, you know, brains or Luxor. If you read the new report by LT, I'm not sure if anyone's seen it. So, yeah, we're top gun. >> All right. Uh, I see that time is flying, so we'll move to the to the last uh section about cost control and margin discipline. Uh, and I wanted to ask a question to you, Alexander. Um there's a view that uh immersion cooling hydrocooling units they um um operate more smoothly uh suffer less from the temperature swings and therefore have less repairs. However, such a buildout is more expensive. I see now that the machines are cheap like many are still uh opting for air cooled uh a lower capex and then a quicker return on investment. How do you uh look at this trade-off between maybe investing a bit more on the on the front end and have a more stable running operation versus uh uh optimizing for the capex on the on the front end and and maybe having to deal with some more issues down the line. >> So yeah, it's true. So for my opinion also the hydra cooling is more efficient and more stable than air cooling. But uh we try to choose it. It depends on a few things. The first one is a region. So for example on here on US market hydrocooling is doing well and uh it's a lot of companies who maintains it. So and also the horizontal of planning over here it's little bit more. So you can in the in the loan or in the middle perspective it will be better but when we working on the emerging markets for example it can be Russia Africa or something like this over there it's much more easy to operate with air cooling because it's easy to maintain things easy to sell and uh also there is not so a lot of uh experience how to operate with the hydra cooling. So for our opinion, it depends on the planning and on the market you're working in, >> right? Yeah, that makes sense. Um Dan, going to you. Uh one of the most powerful optimization tools for for miners is now to strategically curtail based on pricing signals and also to participate in demand response uh programs. uh how does management software fit in there and and what uh what does it do in order to help uh miners be able to participate uh in these uh strategic containment programs? >> Yeah, absolutely. So with with our management software farm God, we can connect into the APIs from your power provider and custom configure your environment to react exactly how you need it. And you can even do it in tiers. So uh if you're doing a voluntary curtailment based on market power price then you can have you know price hits this tier anything 30 jewels or terraash or worse is going to curtail price drops to this tier you know 25 jewels or terraash. You can custom configure your curtailment and tiers to be profitable where you need as well as in the demand response space where they say you need to relinquish now bring everything down and then our APIs to connect into your building management solutions. PDUs or fans can help reduce some of that parasitic load as well. So when you are curtailed and power prices are high, you don't have to run all of your blinking lights. You don't need all of your minor fans and building fans on idle, which is still consuming power at that higher cost. We can configure that in and and develop the uh integration to drop those costs down as well. >> All right. Uh I I touched briefly already about that uh pivot of Bitcoin miners moving to AI compute. I understood uh that at Interest you've been deploying some uh GPUs uh for uh for AI workloads. Uh Alex, could you uh explain a bit more what what does it look like in in practice and how does that contribute to uh margin optimization? >> Yeah, it's true. So inter launch uh its first AI project. So right now it's in the beta our on our opinion. So right now investors love AI. So they don't love crypto anymore unfortunately. So and for our opinion also we understood that our experience is enough to start AIS it's a little bit more complicated it's another um tiers of data center so it's more complicated more expensive but however this experience can um can give us the chance to do this first steps. So right now it's too easy to to get the results but for my opinion is very perspective and it's can be uh additional way how to increase the incomes uh if this crypto winter will be longer than we than we expect. >> All right, thank you for that. I see that we're running out of time uh heading to the end. So I wanted to conclude uh with this question uh and I think uh yeah we've we've been talking about a lot of options how to optimize your your operations. I think the main message here uh is that there's uh there's opportunity to improve your your efficiency and it doesn't always need to be uh the most costly solution. Um so I would really recommend everybody to uh that runs an operation to to look very critical at what they're uh what they're doing where they can uh can optimize. Uh I want to thank the the speakers. Um thank you very much. You all have a a stand further down the hall. So I would like to invite everybody that's sitting here and wants to uh continue the discussion uh with the panelist uh to to head down there to their stand and uh and have a talk with them. Thank you very much and uh thank you for joining uh and enjoy the rest of the conference. >> Thank you. >> Thanks Nico. >> Thank you. Keep it going. >> Every year, this community comes together to celebrate, to debate, to build what comes next. And every year, the stage gets bigger. Sound money, center stage. So, where do you go to celebrate the next chapter in Bitcoin history? You come home. Nashville, July 2027.