Intel·ligència artificial Criptomonedes Mineria GPU Kaspa RTX 3070 CMP 170HX Cost d’oportunitat Biaix retrospectiu

«Soc un pèssim miner de GPU»: la trampa de mirar enrere i veure només els guanyadors

Red Panda Mining repassa les oportunitats que creu haver perdut amb Kaspa, Pearl i un magatzem ple de GPU. El relat mostra una tensió real entre minar només allò rendible i especular amb monedes noves, però també el biaix retrospectiu: recordem el token que va pujar i oblidem tots els que no van pagar l’electricitat.

En un graner ple de bastidors, targetes i peces que fa anys que no utilitza, Red Panda Mining arriba a una conclusió severa: «soc un pèssim miner de GPU». Ho diu mentre desmunta suports d’unes NVIDIA CMP 210 per reaprofitar-los en unes Tesla V100 destinades als seus servidors d’intel·ligència artificial.

La confessió no neix d’haver minat sempre amb pèrdues. El creador té una tarifa elèctrica baixa, ha operat equips rendibles i diu que encara lloga algunes GPU per a simulació. El que li pesa és el guany que veu ara que hauria pogut obtenir: no haver minat Pearl al començament, haver venut Kaspa massa aviat i haver comprat acceleradors que van quedar aparcats.

És un vídeo honest sobre oportunitat, inventari i cansament. També és un exemple gairebé perfecte de biaix retrospectiu. Quan ja coneixem la moneda que es va revalorar, la decisió de no minar-la sembla un error evident; en el moment original era una aposta entre moltes altres que podien acabar a zero.

El vídeo conté promocions de Terra Compute, Vast.ai, targetes RTX 5090 i una cartera Tangem. Les rendibilitats i els casos que explica són experiències declarades pel creador, no resultats auditats ni una recomanació d’inversió.

Un magatzem de decisions passades

La càmera recorre Octominer X12 amb Radeon RX 470 de 8 GB, CMP 210 que fa més de dos anys que esperen una feina, dos equips amb BC-160 i desenes de peces acumulades. Red Panda explica que sovint compra o rep maquinari amb la intenció de gravar el muntatge.

Crear el vídeo, però, exigeix preparar l’equip, filmar, parlar, editar i publicar. El desig de documentar-ho tot retarda l’ús productiu del material. Algunes targetes no s’arriben a encendre.

Aquest és un cost d’oportunitat més fàcil de mesurar que una criptomoneda que no es va minar. Un actiu immobilitzat:

  • ha consumit capital des del dia de compra;
  • perd valor amb cada generació nova;
  • ocupa espai i requereix manteniment;
  • pot quedar sense controladors o mercat de revenda;
  • no produeix mentre espera el projecte perfecte.

La primera millora operativa no és descobrir el pròxim token. És reduir el temps entre comprar una GPU i decidir si es desplega, es lloga o es ven.

Què significa fer mineria especulativa

La mineria orientada a rendibilitat converteix electricitat i potència de càlcul en la moneda que paga més en aquell moment. L’operador pot vendre les recompenses per cobrir les factures i obtenir un marge.

La mineria especulativa fa una altra aposta. Busca una xarxa nova, quan hi ha poca competència, acumula tokens abans que tinguin liquiditat i espera que una cotització futura compensi el risc. És econòmicament semblant a comprar una opció:

  • la pèrdua màxima inclou electricitat, desgast i temps;
  • molts tokens no arriben a un mercat líquid;
  • un nombre petit pot multiplicar el seu preu;
  • la decisió de vendre continua sent tan important com la de minar.

El vídeo recorda sobretot els casos que van funcionar. Per saber si l’estratègia era bona caldria incloure totes les monedes provades, també les que no van pagar res.

Pearl: el cas que activa el penediment

Segons Red Panda, un conegut va començar a minar Pearl durant les primeres setmanes, quan la xarxa encara tenia poca potència. Amb centenars de GPU de la sèrie RTX 30 hauria acumulat prou monedes per vendre’n aproximadament 30.000 dòlars quan van començar a cotitzar al voltant del nivell que descriu el vídeo.

El creador havia rebut avisos sobre el projecte i els va ignorar. Després d’anys de propostes noves, diu que s’havia tornat cínic: cada mes apareix una moneda que promet una oportunitat extraordinària.

La història no permet calcular una rendibilitat replicable. No coneixem el cost de l’electricitat, els equips utilitzats, les recompenses exactes, la liquiditat de venda ni totes les altres apostes fallides d’aquell miner. La xifra és una anècdota atribuïda a una tercera persona.

També hi ha risc tècnic i de mercat. Una xarxa petita pot canviar l’algoritme, patir una incidència, concentrar el hashrate o no aconseguir una borsa amb prou compradors. Veure el preu màxim d’un gràfic no prova que es pogués vendre tot l’inventari a aquell preu.

Kaspa i la diferència entre minar i conservar

Kaspa és el segon gran «i si...?» del vídeo. Red Panda diu que en va minar amb GPU durant unes setmanes, però va vendre quan encara valia molt poc. Amb el temps, l’augment de preu va convertir en guanyadors espectaculars alguns miners que havien conservat les monedes.

La política monetària oficial de Kaspa ajuda a entendre per què les primeres etapes eren diferents. Després de la fase inicial, la recompensa es redueix suaument cada mes, de manera equivalent a una reducció a la meitat cada any. La documentació del projecte estimava que prop del 95% de l’oferta ja estaria minada el 10 de juliol del 2026.

Una emissió més alta al començament permet acumular més unitats, però això no diu quin preu tindran. El retorn d’un miner de Kaspa va dependre de dues decisions independents:

  1. assignar GPU a la xarxa abans que augmentessin la dificultat i el maquinari especialitzat;
  2. conservar el token durant la revaloració en comptes de vendre per pagar costos.

Dir «hauria d’haver minat més» incorpora retrospectivament també la decisió d’aguantar. Molta gent que va minar aviat va vendre de manera racional per finançar l’operació.

Els encerts petits i els fracassos invisibles

El creador recorda un guany d’uns 3.000 o 4.000 dòlars després de minar Karlsen durant tres dies i vendre quan va arribar a un mercat. També menciona un altre projecte amb alguns milers de dòlars de retorn.

Al costat d’aquests casos hi ha Ravencoin i moltes altres monedes que, segons explica, no van pujar quan ell les minava. Aquesta part és decisiva. Si només es comparen Pearl i Kaspa amb l’estratègia conservadora, la mineria especulativa sembla òbvia. Si s’inclouen tots els projectes sense liquiditat, la distribució s’assembla més a capital risc: moltes pèrdues i pocs resultats grans.

FINRA adverteix que els criptoactius tenen una volatilitat elevada, liquiditat inferior a actius tradicionals i risc significatiu de perdre tota la inversió. En mineria, aquest risc s’afegeix al cost físic de produir els tokens.

Ethereum va canviar tot el mercat de GPU

Durant anys, Ethereum va proporcionar una demanda enorme i relativament líquida per a targetes gràfiques. La fusió del 15 de setembre del 2022 va substituir la prova de treball per prova de participació i va acabar amb la mineria a la xarxa principal.

Ethereum estima que el canvi va reduir el consum energètic aproximadament un 99,95%. Per als miners, també va alliberar una gran quantitat de GPU que va buscar rendiment en xarxes molt més petites.

Aquest excés d’oferta explica part de la frustració que mostra el graner. Una targeta podia ser rendible sota l’economia d’Ethereum i deixar de tenir una càrrega equivalent després de la fusió. Cap decisió individual podia crear prou demanda alternativa per a tot aquell maquinari.

L’algoritme decideix quina GPU és bona

Red Panda havia substituït part de les RTX 3070 per CMP 170HX i altres acceleradors amb memòria HBM2. Aquestes targetes ofereixen un bus de memòria molt ampli i funcionen bé en algoritmes limitats per amplada de banda.

Pearl, segons l’experiència explicada al vídeo, va afavorir operacions que aprofitaven millor els Tensor Cores. Una RTX 3070 disposa de Tensor Cores de tercera generació i un ecosistema de controladors més general. Les targetes especialitzades que el creador conservava no van destacar en aquella càrrega.

No existeix «la millor GPU per minar» sense especificar:

  • algoritme i implementació del miner;
  • consum a la configuració real;
  • memòria, amplada de banda i tipus de nuclis;
  • dificultat i recompensa de la xarxa;
  • valor de revenda i usos alternatius;
  • refrigeració i disponibilitat de peces.

Comprar per a una moneda concreta concentra el risc. Comprar una GPU versàtil pot donar menys rendiment màxim, però més sortides quan el mercat canvia.

De la mineria al lloguer per a IA

El vídeo mostra una transició interessant. Els suports de les CMP 210 serviran per muntar Tesla V100 de 16 GB en servidors d’IA. El creador afirma que un client lloga algunes CMP 170HX per fer simulacions matemàtiques i li genera al voltant d’un dòlar diari per targeta.

La V100 va ser dissenyada per a centres de dades: la versió PCIe combina 16 GB d’HBM2, bus de 4.096 bits, fins a 900 GB/s d’amplada de banda, ECC i 640 Tensor Cores. És una eina més adequada per a HPC i aprenentatge automàtic que moltes targetes exclusives de mineria.

Però passar de cripto a IA no garanteix ocupació. Plataformes com Vast.ai són mercats on cada amfitrió fixa preu i la demanda fluctua. El cost total inclou:

  • electricitat i refrigeració;
  • emmagatzematge i trànsit de xarxa;
  • connexió estable i adreça pública;
  • disponibilitat i reputació de l’amfitrió;
  • compatibilitat CUDA i memòria útil;
  • temps sense clients;
  • administració, seguretat i suport.

Una tarifa anunciada no és el mateix que ingressos durant les 24 hores.

La promoció de Terra Compute exigeix diligència

Enmig del vídeo, Terra Compute presenta un model en què el client compra un servidor, l’empresa l’allotja i busca llogaters corporatius. L’anunci promet que no cobra una quota mensual, que els incentius estan alineats i que el comprador conserva el maquinari.

És una inversió en equip gestionat per un tercer, amb risc de demanda, operador, depreciació i revenda. L’al·lusió a la deducció nord-americana de la secció 179 només pot ser rellevant per a contribuents dels Estats Units que compleixin les condicions; deduir un actiu no recupera tot el seu preu.

Abans de valorar una oferta així caldria revisar contracte, propietat i número de sèrie, assegurança, prioritat de cobrament, percentatge d’ocupació històric, tarifa neta, avaries, terminació i cost d’enviament. El vídeo és un anunci, no aquesta diligència.

Un quadre de comandament millor que el penediment

El creador pot convertir la seva reflexió en una política mesurable:

  1. Fixar un pressupost petit i separat per a mineria especulativa.
  2. Limitar cada projecte per dies o cost elèctric.
  3. Registrar tots els tokens, inclosos els que no arriben a cotitzar.
  4. Definir abans percentatges de venda per recuperar costos.
  5. Mesurar rendiment real al comptador, no només al programa.
  6. Assignar una data de desplegament o venda a cada GPU comprada.
  7. Comparar mineria, lloguer, ús propi i revenda després d’impostos.

Això no elimina la incertesa, però evita reescriure l’estratègia a partir de l’últim guanyador.

No era necessàriament un pèssim miner

Red Panda acaba reconeixent que minar només quan hi havia benefici, aprofitant una electricitat barata, també li ha generat retorn. Potser no va capturar els multiplicadors extraordinaris, però va reduir el risc de pagar energia per tokens sense valor.

El graner ple de hardware sí que revela decisions ajornades i capital infrautilitzat. En canvi, no haver apostat fort per cada moneda nova no és un fracàs per definició.

La bona pregunta no és «quina moneda hauria d’haver minat?». És si una regla decidida abans, aplicada a totes les oportunitats i comptabilitzada amb totes les pèrdues, hauria produït un resultat millor. Sense aquest registre, el gràfic del guanyador sempre fa que el passat sembli molt més fàcil del que va ser.

Contrast i context

Fonts consultades

7 fonts
  1. 01
  2. 02
  3. 03
    Ethereum Foundation The Merge
  4. 04
  5. 05
  6. 06
  7. 07

Font de treball

Transcripció amb marques de temps

10 fragments
Consulta la transcripció
  1. 0:00 , obre el vídeo en una pestanya nova

    Hey guys, Arteam here, hope you're doing well and I hear a squirrel in the ceiling. Sorry. Hope you're all doing well. Have in a really great day. My fish. I miss my fish. It's been, I don't know, it feels like a year since I've seen my fish. I love my fish this thing. It's so beautiful. I seriously have no friends. Okay. I don't have really anything for this video other than just me renting and I'm here in the barn because I need to grab all of these I-O brackets from these CMP 1002 tens that have been sitting on this table for I don't know how long more than two years anyway these brackets I need for the Tesla V116 gigabyte cards that I just recently bought for my AI servers and so I need to remove these from the CMP 1002 tens because apparently these fit right onto those Tesla V100 so I'm just gonna remove I think I need 10 of them, so yeah, 10 of them will be without the brackets. But man, it feels good to be in the barn. Oh. My octominers here. These are the octominer X12s. These are the famous or infamous ones from PCSP. Some of you guys may remember when these were like what? $250 including the RX470 AGB cards that are in them. They're all most of them are still populated with these RX470 AGB cards right here. Man, I'm gonna take them all out or I say that all the time, but I wanna take I'm just gonna donate them to my local

  2. 1:55 , obre el vídeo en una pestanya nova

    PC shop or something because I'm I'm never gonna use them but a few these optimators do have the CMP 1002 tens in them I think one rig and then I have two rigs that are BC 160s because I did buy a bunch and I still have a whole bunch sitting here that I've never used including the CMP 1002 tens which you know what let me put my camera down because I'm just gonna start rambling here and removing the I-O brackets enterprise AI is the biggest money story of the decade, and it's been closed to you until now. Tara Compute builds Enterprise Grade dual epic GPU servers. After three years of R&D at their own data center, Tara Compute is now open for business. Hosted in their own data center at a former US military base in Indiana on a locked utility rate, no marketplace can touch. You, by, we host, no monthly fees, we rent it to Enterprise buyers who pay a premium, and you earn more than the regular platforms currently out there. If we don't rent it, We don't get paid. Write the whole purchase off with section 179 and at the end of your contract, if you're not happy, we ship the server to your door, you own real hardware, real data center, fast deployment, real returns, inquire through a special form link in the video description below. Tara compute infrastructure as a service. Okay, so, this one's one. I have one already, so I'm gonna put this over here. And man, this place this barn has been ravaged by a couple squirrels, the ceiling tiles have been coming down. Anyway, so these CMP 1002 tens, I just, man, you know, I just have to say that I'm a terrible GPU minor and I just, I don't know, I feel like I've wasted...

  3. 3:51 , obre el vídeo en una pestanya nova

    Not that I've wasted you know getting these GPUs on but wasted the lost opportunity of not having the setup in the opportune time of like when I receive them and I'm not gonna sit here and like give excuses or anything but I'm just gonna like just talk out of my head here on you know the things that I've I've probably missed over the years and it's not just regarding these cards the CMP 102 times that I bought here but which I think I got them I don't know two years ago pretty much the same time when I got those those PCS people the Octominer X12, but just like lost opportunity. I'm just gonna say like overall as a, I guess, crypto mining YouTuber, right? I do all this and create content. And one of the biggest hurdles I have as a content creator is when I receive hardware like this, I want to film it. and I want to film me building the rigs and whatnot. But that obviously takes a lot of time, it takes a lot of work and I got a record and talk at the same time and cut out all the umms and all the blurbs and whatever, all the bad stuff that comes with content creation and trying to make a great video, but it's hard, right? It's hard. Again, I'm not trying to make excuses here, but why I'm just like, I'm gonna say I'm a terrible GPU minor. Just comes in terms of just the opportunities in which I probably could have made more money if I just focused on maybe trying to spec mine a coin or trying to mine the coin before I even, oh, we hold on. I'm missing one screw, hold on, one more screw. I could have, I could have specmined some coin and made more money doing it that way and you know I'm gonna like attribute to this this type of opportunity like specmining opportunity to Just recently someone told me they were you know specmining pearl okay Pearl just came out I think like not too long ago like two months ago and

  4. 5:46 , obre el vídeo en una pestanya nova

    They told me that they were able to know about this coin right away and they were able to pretty much, you know, mine it when nobody was on the network. And there was a good couple of weeks, you know, Pearl was out and nobody was really mining it. I think mine is the... minus the devs, I think, or the team. I'm not sure because there's a lot of rental, a hash rate rental for Pearl, right? Because Pearl requires, or the algorithm at the time, I guess required tensor cores, which, you know, the regular gaming GPUs, AI, AI, BGPUs were really good at that, not the CMP cards, like my 178 checks is, and that's a whole other thing I'm going to talk about in a second. But, you know, just, Pre-mining, you know, that coin, this guy, pre-mined like 30 grand. When it went on exchange and when it went on, you know, when it was like over $1.50 or $1.50, the guy sold, like he said he made like 30 grand. From just mining over like two weeks or so. And he, he mined a bunch of pearl and he had a bunch of, you know, octamineers and like, 30 series cards he didn't have a lot of 40 series stuff but he's still he was just mining with 30 series cards that he had at the time and he he took the chance right he took the risk and the the gamble you know trying to spec mine this coin and all of a sudden it went on exchange and you guys you know you guys know the rest of history the coin exploded the emissions of of Pearl was like, what, two, four, five, two million or something. Two to three million, I think three million at its peak or something. But it was a crazy coin that just came up recently, I don't know where and had all this emission for GPU miners. And so, yeah, just crazy, right? Like stuff like that, like to be honest, again, that goes back to, I'm just a terrible GPU miner because I was actually told about Pearl mining when it came out. And I gave no thought to it, I gave no.

  5. 7:42 , obre el vídeo en una pestanya nova

    I didn't believe in anything because I've spent mine a lot before and you know, I just I was just how do I say it this heartened cynical about it and I just never I just never bothered and you know it's not like I was gatekeeping it like I could have made a video about pearl mining a long time ago before it was all hyped up But because some you know that I'm still get a few people who tell me like hey Red Man, you should mine this coin it just came out or you know I get I get told that like every other month or every month when something new comes out I'm almost told like hey Red Man this new coin came out you should GPU mine it I just never I just always skipped it. I always over the past. I'm gonna say over the past couple years. Ever since like 2021 or something, I can remember like, you don't know if for anyone that's watching like GPU mining Caspa, right? Who GPU mine Caspa early? I mean, I did for like maybe a couple weeks, but like I didn't do it that much in a sense. Like, and then I sold the Caspa when it was like, when it was like nothing, when it was still GPU-minable and good, right? I are not good, but like, you know, You could accumulate a lot of Caspar if your GPU-minin held it until it went on exchanges and went up to like 17 cents or something, then you could have made out like a bandit. Then you would have been like, that's the best play in terms of GPU mining for those that GPU-minin Caspar. So again, it goes back to like, I'm a terrible GPU miner. I've never, I've never, like, you know, People always preach, you know, spec mine or mine this coin before goes on exchange and I never did that like I know I don't think I have many How do I say it? I don't think I have many wins in terms of coins that I you know GPU mind and when it went up or I went to on exchange and it went up in price. I never had many chances of selling, you know. The only time I could think in my recollection that I ever made a video about in which I specmined a coin was Carlson. I GPU mine Carlson with like a lot of my rigs for three days and I accumulated enough to sell about three or $4,000 worth of Carlson when I went on exchange. And I sold it.

  6. 9:37 , obre el vídeo en una pestanya nova

    and I made three grand in like three days of mining it. So, you know, I felt that rush. I felt that, um, how do I say it? Yeah, just, you know, GPU or spec mining something early and selling it when it went on exchange. And I felt that rush. It was great. But after that, I never really, you know, because I've done it before on many other coins, actually, uh... I lied. There's one more I did was on Lee Fiumm. I did GPUMind that and was able to make maybe a 2 or 3,000 off that but nothing crazy, right? Unlike the guy who I just talked to, who told me about he was pro mining with a, you know, a couple hundred, a couple hundred GPUs and he was accumulating Pearl and it worked out for him, right? So, you know, it's things like that where... There's many times that has happened and I missed the pearl pump like I missed the spec mining pearl early which I should have done and I could have made out like a bandit, but I didn't right so I'm just yeah just It just goes back to these GPUs sitting on the table. I've never had these on these have never been on like I bought I think 40 48 42 of them 12 of them are already in the next 12 and you know and like these CMP cards they were great for a lot of these are these are really great for the memory depending coins and you know you all remember like when like core dependent coins like Caspa like they were FPGA and then or like the other core, depending on coins, next, there's a whole bunch that went, eventually went asik. You guys know? And I thought like my whole reasoning as to going with these types of GPUs, like the 178 Jecks is what you all know, I still majority have.

  7. 11:33 , obre el vídeo en una pestanya nova

    And thankfully I got one guy that's renting them because they're actually pretty good on FP16 performance so he's doing some math simulation on those GPUs which is pretty good so I'm making about like a buck a day per 178 checks which is nice so still making money on those thankfully, but you know in terms of GPU mining since Pearl came out the tensor cores on the 178 checks and you know the CMP cards were never good It's they're not good for core dependent stuff only for memory because of these GPUs have the HBM2 memory and so the memory bandwidth on these are just like 4,096 bitbuss which is insane right? So that's what made these GPUs really good on like memory dependent coins like or a little bit of core stuff but like I think Capau is pretty strong on these right Capau? E-Tash right E-Tash these could get like I don't know 60 to 100 mega hash or something And these are really good, right? And so there weren't many coins, I guess, after, you know, Ethereum really that were truly strong for these GPUs. But ever since Pearl came out, the CMP cards because they take advantage of the tensor cores. These never, these never, became profitable or useful or yeah profitable I guess and Didn't didn't have very high hash rate right compared to like an RTX 3070 so hindsight 2020 I should have kept all my 3070s instead of replacing them for 178 taxes but actually pearls kind of been coming down in price, but it's actually still good I see a 3070 still making like 75 cents a day after my power costs so it's actually still decent but you know, it didn't didn't last right and so yeah but anyway yeah I just want to say like I'm a terrible GPU miner guys like I I never truly took advantage of you know trying to speck mine something and then ultimately you know see some coin going exchange and then I make like a million dollars or at least six figures or something from GPU mining it I've never had that

  8. 13:28 , obre el vídeo en una pestanya nova

    You know, I've never had a win like that before. Let me know down below. If you guys have ever had a win like that, if you guys watched this far, I'll give away two pedaache for two hours. Just comment down below. If you ever, like tell me what your guys' biggest GPU mining spec, mine coin, win that you guys have had or pre-asic, you know, you mine something before the A6 came out, then the coin pumped because of the A6 and then you made a bunch of money. Let me know. I like let me know if you guys have a... Let me know if you guys have ever Which coin that you're able to make a lot of money on I know that a lot of people for a fact made a lot of money Gp mining Caspa and so yeah, and then they held it until you know, maybe when it went to four cents or two cents I know some of attack was a lot of people talk about him and when he sold and all that kind of stuff I got nothing against him or anyone who sold early or you know, you just, you know, no one knows, no one ever knows. But if they made like a 20x or 100x or whatever type of x, multiplier from something of a coin for a much higher price, good for them, right? Good, good on those guys, right? And so it's, oh man, this one is freaking stuck or rusted. Okay, yeah, there we go. I just gotta, yeah, this freaking, Screw is not coming off Yeah, anyways guys. Yeah coming down below if you watch this far Thank you all for watching if you watch this far I know who watches and who truly watches my videos day in and day I really appreciate all you guys But anyway guys. I'm just ranting right and Just kind of just talking about you know some of the wins and losses that I've majority losses that I've you know a GPU mind Raven coin

  9. 15:24 , obre el vídeo en una pestanya nova

    in the worst opportune time, it never went up and a bunch of other coins are of GPU mind and they never went up. So ultimately I feel like it just feels like I'm at a massive loss with all the hardware that I've bought over the years and it's just been like, you know, okay, well it's not a loss but because of my electrical rate, instead my opportunity for myself is if I have hardware on that's mining, it's only during profitability. So instead of spec mining waiting for it going or exchange I'm able to mine especially the past couple years GPU mining stuff that is profitable and so for me I guess my ROI for some of this stuff or maybe all of it already is that I've been able to profit from having cheaper electrical rate than most people and then you know Making money that way right? What do you guys what do you guys think let me let me know? Right? Let me know if you think that's that's happened with a lot of you guys as well, you know, GPU mining for profit which I think, you know, I think a lot of us did and just make sure here I have eight. I think that's all I need for the eight V100s. Yeah, perfect. Perfect. Okay guys, I think that's it. I'm going to call it for this video. Let me know your thoughts and yeah, I just just kind of reflecting just a Remember, in this thing of the past, I can't believe I have all these here. I bet you one day though, some e-tash coin is gonna come out, or some memory dependent coin, GPU coins gonna come out, and these CMP 1002 Tans on my 178 juxes are gonna be good again.

  10. 17:19 , obre el vídeo en una pestanya nova

    and maybe even the BC ones, 60s that are sitting on this box right here. Maybe these could be good. Who knows? Okay, my friends, that's it for this video. Oh, there's just a filler video basically, but just let me know your guys, the thoughts, what I just talked about. And if you guys relate to it in some way, I'm sure you guys do. But thanks for watching. Again, comment down below. Two pedagy hours from Brain's Hashpower. I'll see you all in the next video. Have a good one. Peace out man. It's it's been a long time since I've been in this barn God I got to throw out all these I got just barn is just packed I got to throw out all this crap. I have so oh my god my mining frame graveyard It's just oh my first mining frames from 2017 back in the beginnings of Ethereum mining right there. Oh man good times my 24 inch exhaust fan there Okay anyways see you guys peace Peace peace peace God. I've wasted so much money and time Still storing your crypto on exchanges That's risky. This is the Red Panda Tangerine card. Simple, secure, and built for real self custody. No seed phrases, no complicated setups, just tap your card, manage your crypto, and your in-control. Instantly, secure chip technology keeps your assets protected, even offline. I've tested a lot of hardware wallets, and this is by far one of the easiest ways to get started. If you want true ownership of your crypto, check out the Red Panda Tangerine card below. Simple, secure, self-castody used code RPM for another 10% off. Link in the video description below.