Chia News Livestream Edition – GPU Plotting On Pascal Cards, Math, Hard Forks CHIP + More! | CoinMarketBag

admin

An empty audience because I do that usually every single time and yeah I hope you are having a good one here’s one for you and we’re gonna do some plotting tonight we’re gonna go through some math I did some math on the plot filter uh we’re gonna look at some Implications of the…

An empty audience because I do that usually every single time and yeah I hope you are having a good one here’s one for you and we’re gonna do some plotting tonight we’re gonna go through some math I did some math on the plot filter uh we’re gonna look at some Implications of the plan chip what that means for disk seeking activity we’re also going to get the P4 plotting this is the Tesla P4 this is a passively cooled cheap little card uh it’s actually probably a really good card for farming based upon some information that I just got if you are going to be under a petabyte or something like that uh talking about petabytes look at all those discs so we’ve got some storage stuff that has just uh been filmed this morning and the wrap up of that is going to be the remainder of this weekend so This will be the only video that I’m doing the remainder of the weekend but there will be a bunch of high performance storage topics catered to you in Chia and I promise we are not going to get into anything insane like looster maybe maybe later definitely not up front but I hope Everybody’s having a good weekend and yeah so I mean the price of chi has gone kind of crazy and that has me excited like it is insanely exciting because we are hitting 45.somebody said we Wicked up to like 50 today now I don’t see it on this chart I don’t know If we would be looking at that on a different chart or something maybe the time scale that I have here isn’t right uh I don’t know but this is this is exciting to see Chia here and do you want to know why I’ve got news for you you probably have never heard Chia is now as profitable per ten thousand dollars that you have riding on it as ethereum was right before the merge that is right that’s insane and I mean I I don’t think most people have been like keeping tons and tons of statistics on this but uh let me see here digital Spaceport I have been keeping statistics on this for almost a year now because I knew that there was going to be some changes and we have crossed the 20 dollar per 10 000 threshold on your Chia Farm if you look here our nearest competitor down here is what Raven at three dollars And 45 cents in the hole on that same ten thousand dollars that’s probably not factoring in some things that I should actually probably redo this and Factor them in I’m just being consistent this is the merge and you can see we’re at about nine dollars and 37 cents look at Where we’re at today 20.47 and that’s per day earnings on your Chia operation that is insane that is like that is wildly insane the cat I see you talking about the cat the cat could end up being right before we’re up before we know it the cat could be right and that last Prediction of the cat I hope is damn right because God damn that was a huge prediction so yeah uh let me catch up really quick here on the chat and see what’s going on Austin uh loving him some hard forks I think he’s got some big did over there that he might be Loving on also I don’t know slow stuff I’m more this is John uh she likes John uh his did his Fork can only get so hard I I would imagine Daryl uh hard spoon type of person uh uh TJ uh nice to see you here tonight also I see you’ve played uh knifey Spoonie before wow uh the chat’s getting dirty here you guys have fun with each other by the way if you’re in Discord feel free to DM each other see what kind of uh action you can get on the side I don’t know uh Gordon hello how are you doing tonight Uh just found out about the GPU plotting how much RAM do I need for level 5 compression plotting good news for you sir as far as level five compression plotting you probably can just get by with like almost anything that you’ve got out there unless you want to like When Shia comes out with their plotter you can use any of this with the CPU except for above C7 with the Giga horse plotter and at that point you need to have you know 256 gigs of RAM if you want to do k-32s we’re gonna talk Tonight like about should you do k-32s I did some math around k32s uh so that might help us it might confuse us more uh John check my math at the same time in real time and call me out if I’m wrong on any of this but I think that What we’re seeing with the planned hard fork and again we’re going to step through this because this is kind of chia news and stuff but if you look at where we’re at with the hard Fork we’re at the chip has been proposed and everybody should be if You’re a farmer commenting on this sign up for an account on GitHub let me show you where you can actually find this information out here this is on GitHub and I’m going to take this link and toss it into the chat here really quick so that everybody bookmark this check it Frequently that way you can stay up to date on what’s going on with this you can comment on this as well make sure you do let them know what you’re actually interested in what you think represents you and your farm I would say that we are at a Crux in time where it’s Going to be hard for me to imagine the security of chia being super viable long term if they don’t make changes at this point in time and I I know there’s been a lot of people talking about burst coin and the similarities between what’s happening now and first first was a Show on many other levels it was literally Rife with scammers and they really didn’t have a ton of development Talent like the team that made it kind of just like flew out the coop and so I mean there was a lot of problems that Chia doesn’t have it’s not something That you can really draw an analogy yes they did have an attack on theirs and yes it did end up being something that after the fact was already going on was discovered there’s a big difference there this is talking about doing something proactively before that kind of show happens and so that is Really what we need also the kind of things that can happen here are not I don’t even know if it’s fair to say they’re malicious it’s just it would if there was true just we’re changing it to just Computing on the fly with GPU power That’s all we need to do all of this then it would essentially be proof of work and so that is the effort to you know not have it become that that she is undertaking here and so that we is why we see the proposals that we see from Chia about things like the plot filter and cutting the plot filter and the impacts that will have drastically increase the amount of work that would have to be done on the other side making it pretty much infeasible for somebody to use gpus to effectively Brute Force Against that and still be able to solve on the fly so I think there is good reasons they’re doing the things they do again even if you you know solve all of the signage points and submit and everything like even if you do that you still are not going to necessarily win Every single time so it’s not like an attack where you’re going to steal everything but it would be a very greedy situation and not necessarily what’s planned and you know proof of space and time the consensus itself is something that is being importantly considered what happens and I can’t imagine this is Anything but a Crux period for the team at Chia to be actually going through this kind of stuff and doing this kind of math and looking at what they have to do to be up to speed so this doesn’t become catastrophe which I think that’s what we’re seeing as responsible planning From the team at Chia at this moment in time towards those things so let’s take a look here really quick at uh this and just you know talk about some of the math so maybe you know you’re going to read this there’s a couple things that I Think you should read here and so one of the biggest things is about the specifics of a 20 terabyte hard drive example this housing 183 k32s okay so I did some math here at the different proposed filter sizes so we’re at 512 right now and so that is essentially One check per 2.8 signage points that’ll happen and so at 256 we’re at one check per 1.4 signage points at 128 we’re at one check per point seven signage points so at that point you’re going to be checking the same disk like twice at one check per at 64.the plot size Filter gets down to 64.one check per .35 signage points so that’s about 2.85 checks per at 32 the filter size you’re looking at one check per 1.75 signage points so you’re going to have somewhere around 5.7 checks per and that is on each disk okay and so there’s Roughly 50 to 70 milliseconds of activity per check and this is something of course because you have a hard drive that is going to be operating essentially uh like sequentially so check check check check so your hard drive needs to kind of spool those and so if you’re looking at 32 for the Filter size that’s 5.7 ish and so on the low side you’re looking at about 285 milliseconds and on the high side you would be looking at about 400 milliseconds for the checks on that drive and that would be happening of course within that five second window For you to submit your proof so I don’t see any problem here now if you start going Beyond this because you can see the rate that this is growing at if it was to go to 16 for instance then okay maybe there’s some different things There and I would note that they seem to have planned this at a timeline that stops at like 20 30 because we probably do have like an SSD flipping of hard drives some time-ish around then when SSD technology is just a prevalent technology and that definitely changes The i o characteristics of drives is even if we just use the same SSD Tech we’ve got today which I’m sure it’ll be way better in 2030 that’s going to drastically change the i o care characteristics of drives and so I don’t think that it’s a problem to look at a 20 terabyte SSD and being like hey you need to run quite a few checks that’s a drastically different proposal than a mechanical hard drive and so it looks like their proposal takes us up to about 400 milliseconds in 2030 for your 20 terabyte mechanical hard drive to be Doing checks by 2030 that mechanical hard drive uh you know it’ll it’ll have seen a very long life I actually expect 20 terabyte hard drives could totally make it to like that time frame so I think that’s actually being pretty realistic uh on on storage jm’s part Here for the proposal and I think that the math makes sense now is it aggressive I don’t know if it’s aggressive or not uh does it cause extra disc activity the little bit like a little bit now once you start actually on a 20 terabyte disk needing to perform a you know check Every single time then I I think you’re going to be looking at your idle States being altered the actual wattage that’s used for use utilizing lookups on hard drives probably increases a little bit as a result of that is it a tremendous increase I don’t think so this is still A relatively light workload and so if you think about it in this terms if you look at like a ZFS array or something like that you’ve got like they’re all spinning all spinning all like so it’s drastically different so every activity is like there there’s things happening So this is vastly different than what you might be thinking of is like oh this is way more electricity it’s still going to have a lot of periods of idleness from what I can see and this is actually why I am actually kind of in favor of this at this point In time I think it makes sense uh I think it protects the integrity and the security of chia’s blockchain and the fairness of earnings potential by Farmers is it a pain in the ass yes let’s take a look at something that uh somebody pointed out which I thought was Pretty interesting and uh this is a poll and it was basically around when will you reply and how and I mean I think I am going to go with I prefer gigahorse plotter by Mad Max and I’ll wait two to three months longer if I have to while everyone else replots the The reality is I I think that it’s going to be not released at the same time these two texts but I’m gonna vote for this one right now and we’ll see and I’m really just a selfish plotter at the end of the day so if the the reality most Likely shapes up to me to be whatever’s out first is what I will use and whatever offers the best return on investment I think that you know people that are following this channel that’s something I’ll be talking about and we’ll be I’m sure looking at spreadsheets these spreadsheets are Actually very difficult to put together I’ve already started trying to model some of this out and I’m like oh there’s like more variables that I’m used to in this so based on case eyes based on some other things there are more variables so I don’t have a sheet ready yet that Looks at what I project for profitability but the other thing is it’s all theoretical at this point we we don’t have the Chia version of from Mad Max yet we don’t have the Chia official version yet as a result of not having those everything we can do is basically Modeling and we’re going to actually go through some modeling tonight after we do some plotting on these two different gpus so let me catch up with the chat real quick uh find out where everybody’s at here oh wow yeah there’s quite a bit of chat activity there um I bet an fpga could plot faster than 28 seconds well in so I mean there’s tremendous cost with fpga development with Asic development and literally if the if there is a variable moving like the plot filter it would become obsoleted and there are other there are other other mitigating factors that Shia Could there’s lovers to pull more than just the plot filter the plot filter is a big one like this is this is a really big one uh there’s case eyes there’s all sorts of other things and that kind of obsoletes some of those texts so I think Chia is definitely as far as I I’ve never heard them say this per say that they’re Asic resistant but I I mean everything Chia talks about is Asic resistance so I can’t imagine them being like let’s go along with you know Asics that are plotting like I don’t think so That doesn’t make sense to me because that would essentially be a proof of work by the time you get to the cost investment for it to make sense for an Asic hard drives would be out of the equation at that point and it would literally be just the grinding Uh yeah glad you guys can hear me wow I really ranted a long time without checking to make sure you guys can hear me I am glad that I checked that out uh you only see a great Egyptian in front of this pyramid oh yeah man uh EPO what’s up man glad You’re here tonight the price of big did I have no idea uh you are really excited about this big dig cat out there uh Chia cats again one of the things that we haven’t seen tons and tons of activity in but we recently have like a big did Project out there which I think that’s just some interesting stuff I have no idea what’s gonna happen with it but it seems to be air dropping out there and being given away and all sorts of other stuff if you have stuff and it did and you’ve grown a big did again your Decentralized identity on Chia and you’ve moved a bunch of crap to it if you’ve done that which is a Time intensive process in and of itself then you probably got airdrops some big did I don’t know what to say happens next but we’ll just all have to wait and see uh Doesn’t it Factor compressed plots increase profitability okay that I mean yes okay so like let’s look back at what we just were looking at here like this is like I’m not planning on replotting so I think I was saying I think between 40 and 60 of people are Planning on not replying so we’ve got 97 votes 41 or it’s roughly 40 here that are saying hey I’m not planning on replotting so if you do reply this looks like yes it should actually give you some pretty damn decent returns so I mean I’m planning on replotting I don’t See where people are thinking they’re not going to reply unless they have incredibly High electricity costs or if they a ton of people have just forgot about it I mean these are the people that are cognizant and on our Chia which I mean okay so they’re people on our Chia so there’s That also but like these are people that are actually thinking about champ there’s a ton of people that are like I’ve got a Chia farm that I put up and I literally don’t even remember it and it’s setting in a corner or a closet or something and that thing’s just going to Keep going who knows what version it’s on there’s a ton of people like that also um there’s a lot of people that are just gonna be like eh whatever I’ll just let it ride I don’t care and they’re not gonna comment so if 40 currently on that Poll of people who are you know at least semi-interested and up to date in Shia are saying that they’re not going to I mean it makes it it makes sense to me to think that most of the people out there should be thinking if you’re watching if you’re Paying attention if you have the capability and I’m going to show you a card tonight that will give you that capability for like under 100 bucks that yeah I’m gonna reply keep going here and catching up Ron can I plot low level C with a so I mean the Low level C stuff yeah I mean I think you can plot that right now with gigahorse uh you can you know use CPU plotter for it with but it’s not Chia like it’s not the Chia version like the plotter out there right now is MMX so Yes and I think you should plot some MMX to test it out because you’re not going to know what you’re doing unless you’re literally able to run and you’re not going to be able to hit the ground running which is time and money and I think time in is going to be an Optimizer here like optimize your time so start playing with this Tech now today because if you don’t you’re gonna be surprised with you know scrambling at the last moment I’m gonna try to keep everybody up to date produce guides that are helpful around how to use it so that people can See this is a good build this is a good pathway forward this is you know how I can move the plot files Fast Max I know on his side has created a plot mover that greatly simplifies because you’re creating plots now so fast if you have a Mid-range GPU that I mean moving things to disk is hard and so Max has really helped out with that and I think that’s going to help a lot of people that have a bunch of disks and they don’t really know exactly how to move plots fast foreign Taking a drink every time he says hard Fork hard Fork my friends hard Fork oh yeah I did forget I can’t be super foul-mouthed because uh YouTube actually does Doc money well they don’t dock money but they don’t show ads and stuff like that EPO what are you saying there I have so Many questions that uh you stay up and get all the answers one question though it’s about time to do our full reset on the OS drives on the farm there’s a new Jammy jellyfish okay for the full node and Harvesters or should I just stick with 20.like I think you Should go with 22 like go with Ubuntu 22.you don’t got to rush it you’re not going to gain like that much here I mean you’re gonna maybe gain some kernel stuff that’s not important so I wouldn’t worry about doing it it’s not going to be a priority in my opinion for A little while but at some point you should when you do a refresh go to 22 for sure absolutely I need to chart with all this stuff uh you know I think I can probably work on getting a chart together Bazinga thanks for tossing me the 50 bucks man always Appreciate that uh one of them what was it you said one of the main burst coin persons turned out to be a scammer spent time in jail and also a drug dealer I mean that’s like this is what I would expect from that show that happened over there because that was literally Like I mean there was a kiss granted there was a casino and I’m like I I if there was a Chia Casino I would spend some xch there like online I would totally expense Maxi Asia there so maybe I shouldn’t judge too hard uh SATA hard Drives I’ve got a few and by a few I mean five SATA hard drives that will be listed here on the website shortly and I know prices have increased my prices have increased and so I don’t know what to say I was told about the price increase a fair amount Of time ahead and I tried to give everybody a heads up I was really glad over the holidays that I was able to get such like a massive amount of hard drives moved at like really good prices but New Year they apparently kind of adjust pricing to check out you know What they can get you know as far as pricing in the New Year could go down could stay where it is could go up I hope not but we’ll just have to wait and see basically whatever happens here Ron Watkins PCI 3 bottleneck yet you know if you’ve got pcie4 I I really want to get a pcie4 system I mean I’ve been looking at some epic stuff that I really need to get my hands on and you know 900 bucks for Ram yeah sounds really high to me even for ddr4 I don’t think that it’s going to cost you that much um There’s some chatter about different compression levels and stuff glad you guys are having a fun time talking to each other do that also for farming at lower C values a Raspberry Pi is still good for a moderately sized Farm yeah I mean I don’t think anything’s going to be Obsoleted here like that’s the reality is even something that’s not yeah I mean the biggest Improvement you get is that first step into the Hellman attack against your plots that is the biggest first you know gain that you can get in any of this and that’s going to be with Either Chia official or with mad Max’s solution that first compression level is where you’re going to see the biggest boom save of space and so even if you just did the lowest level that’s still a significant savings uh you know I don’t know evg miners they Might not have an easy way but they might have a program or something where it’s like send in your hard drive and we’ll fix it or you know re-plot it for you maybe they charge a fee for it I don’t know maybe the they have software That they push out that if somebody has a GPU they can plug it in and it’ll go off so I don’t know how that’s going to work for those but I think also the other thing is you can replot your own drives uh so there probably will be a Lot of people that’ll re-plot their own drives that have them now I don’t think it precludes people on evg from re-plotting it’s just how the software is going to work that I don’t know about so that’s a good question though that’s definitely a good question uh I will be reporting bazingus Has OneDrive at a time this way my 1.6 Pepi byte Farm will stay up while replotting that sir is exactly what I think is the best strategy possible because there are going to be a lot of people I think that are going to do silly things like I I you know maybe they’re like oh I’m gonna you know remove two drives maybe I’m gonna remove 10 drives like maybe there will be people out there that just they won’t know what they’re doing at all and they will accidentally delete stuff I think there Could be a huge like there could be people that get mixed up about witch plots which I mean there could be all sorts of things so I mean there is probably a lot of errors that could happen during the re-plotting process I think keeping effective net space you’re like your own personal effectivenet Space really high during that time period is like a killer idea because that is probably you know there could be variables you know like oh what if it dips really big when one of these plotters comes out and is released uh that would be an exciting time to be up If a bunch of people dumb stuff and that see that I did it again if a bunch of people make a dumb mistake and then you know take their space offline so that’s one thing that I would say I think that’s a good idea go slow uh you Might be surprised at how fast this thing’s gonna go though so I think one drive you might consider doing a couple of drives I think four drives with a 3080 or a 30 60 is going to be about what you’re going to be like able to sequentially spool off to so I think Four drives would not be a bad plan uh three drives to four drives somewhere in there will be usually able to handle the max rights that are going to be coming off of your buffer pool so I would consider that eggs have increased I I’m out of eggs at The house we have some pizza though so I’m gonna eat some pizza along here while I’m drinking my beer at some point and yeah we’re out of bags here uh OG bursting was crack cracking BTC wallets yeah you know people were like taking other people’s computational power with That whole thing like it was seriously a show it was seriously a bad thing that was going on there uh I think I need to stick to C1 you know C1 ain’t bad demonetized not I just need to figure out what Hardware is required for each level before I Decide see one for I mean so this is the real thing is like people profiling their Hardware is happening right now on MMX so join the MMX Discord the Mad Max Discord he’s got like channels they’ve got spreadsheets they’re keeping information about this MMX is not like a 100 Chia but you know it has an analogous properties so I think that that is something that you can actually look at and say there’s some comparables here so I would do that all right let’s get to some plotting I know you guys are here like for some plotting so let me Get the uh terminal window featured here that way I can eat some pizza while we’re doing this okay so the first one that I’m going to do here is g0 so this is a new variable that I learned about and certainly if you are in the uh let’s Practice let’s get this right uh kind of mode then you should be looking at the readme’s on each one of these and so if you go to the Cuda plotter here you can see all the usage uh stuff here so all of the things you need for uh Putting in your variables there and you can see here we now have 6.1 that’s Pascal support for plotting so yeah that’s really good and again don’t forget uh you don’t want to be uh utilizing a um oh where is it here uh you don’t want to Be utilizing a uh the first plot for your like statistics you want to be utilizing the second plot for what your actual statistics are going to look like for your plot speeds okay so I’ve got one quick queued up here this is uh two plots coming off onto our eight SSD Array and that SSD array is going to be housing them and yeah C7 that is you know GPU only compression level there personally I think C8 is starting to look kind of attractive but C7 is definitely very attractive I think that k33 is starting to sound pretty decent To me also we’ll just have to see what happens but this should be the let me go over here to NV top yeah we should be using here the uh Nvidia 3080 TI this is all happening in the uh dell t620 over there and that puppy is just a wonderful wonderful Machine for doing this kind of workload also today I got this is exciting a 16 Lane pcie 3 ribbon cable it should hopefully be able to negotiate and support that pcie 3 standard so I can put the 3080 TI eventually into one of the r 930s that can house I’ve got Two terabytes of ram so I mean I’ve got two terabytes of ddr4 that I can toss in there Isn’t that cool if we can plot some like k-35s on that like I don’t even know if that’s farmable uh but maybe Max is interested in finding that Out I’ll have to ask him if he’s interested in finding that out and maybe there’s a chance that we could have some sort of a plot around that that we could see some interesting information I don’t know but yeah this is just it’s so fun to Watch is it because it just goes so fast I mean it just goes so fast 50 minutes that is a expensive competition I think I got it way cheaper on the first competition I really don’t even know where to get eggs our Sam’s Club is just out of eggs like how does a Sam’s Club go out of eggs this just doesn’t even make sense Ron you don’t need that much so I’m showing you literally I’m not gonna say a 3080 TI’s top end but it’s good uh and it’s on pcie3 here and I’ve got five 12 gigs of ram in There which you only need 256.

I’ve got 512 uh and I mean I think that we’re going to see some impressive times I know what these times already look like and so we’re going to look at that that’s kind of up here like this is a top end plot this is what it Looks like then we’re going to look at like the Tesla P4 and this thing is literally a passively cooled 75 watt card so you don’t there’s no power pins for it you just slap that sucker in and that’s it uh and we’re gonna look at the Plots on on that also so I mean there’s two differences there that are I think pretty big and so we’ll get to see a high side and a low side what you’re gonna find is that the like Delta between the 3080 TI and the like Tesla P4 is surprisingly Small I mean if you’ve been following along in the Discord don’t spoil it by typing it in because I don’t think Ron’s been following along in the Discord uh but yeah it’s not that big of a Delta difference and I mean you’re talking about a card that I paid 90 for Versus a card that I fomo’d probably into at the wrong time it’s been a little bit more than 90 dollars so yeah definitely definitely overpaid for the uh 3080 TI given today’s prices for the 3080 tis yeah it’s pcie3 the Dell t620 is a uh Ivy bridge that I’ve got in there it’s Got dual 2697 V2s you don’t need dual 2697 V2s either like if I do a h-top here in one of these I guess I’ll uh do an H top here this is what the CPU profile looks like so I think like you could go with a much lower powered CPU than I’m using here uh and be okay so I you don’t need a Powerhouse with a CPU and this is two sockets also so you definitely don’t need two sockets and I mean if you look here this is not a lot of load that’s being used on this But yeah it is fun to watch this crank along here and you can see if you’re if you’re new to like watching NV top like it’ll tell you the generation that it’s negotiating at here it’ll tell you the transaction speeds that it’s getting up here in the upper left and upper right And like that’s fast that’s really exciting to watch you’ll see how much memory is being utilized down here and you’ll see how much of the GPU is being loaded down here and uh so one thing that you can do here also is uh you can go to this and you can see your Performance State and so I see that this is performance state two I would love to be able to figure out how to effectively Force this on Linux into po I don’t know how to do that I’ve been trying to figure it out and it does change every Now and then and it’ll kind of variably go around but this is not the highest performance state that this card should or could be able to go to so figuring that out is one thing that’s on my to-do list if you know how like like you know Hit me up in the Discord Twitter at gospaceport uh toss in links over there in the chat like that would be awesome because I think I’m least I think I’m losing out on performance I don’t know how much performance but I think I’m losing out on at least some performance The reality is though that first plot is already done the first plot is your slop plot and so we see that that’s 267 seconds this next plot is going to be even faster it’ll be smoking off of there in like no time and so yeah I mean you’re looking at 4.5-ish minutes that’s Freaking fast by any standard like dealing with the plots coming off is going to be hard and Max knows this and so that’s why he has created his tool for handling them I don’t know exactly what she has got that’s going to handle them but I hear There is something that has me excited so what do you I mean while we’re waiting for this plot to cook off and finish here which will literally be just a few more seconds sound off in the chat and let me know what you guys think hard Fork soft Fork I mean are you guys excited about this does this like scare you I mean I’m not really I’m not scared Vibes on this I mean I think this is actually pretty exciting I think we’re gonna see some amazing stuff and I think a lot of people are not going to reply and I Think like us that do are gonna stand to make more as a result of it and I’m albeit maybe somewhat not happy about it but I am a greedy person as a GPU Miner in the past as a Shia farmer now like it’s fine to just be honest with Yourself and be like hey you know I’m in cryptocurrency because I want to make money like right kind of what everybody should be saying I would think 157 seconds so 2.6 minutes is what this plot just came off there at So that’s fast right I mean that is like Very fast uh and that is on a pcie 3 system this only used 256 gigs of RAM for this this was a k32 not a k-33 plot this was the latest binaries that I just pulled down earlier today if you are going to be trying out Pascal Era cards A 1070 uh if you have a P4 or a 1080 or 1080 TI make sure you pull down the latest binaries I forgot to do that earlier and some testing and it was sloppy sloppy slow uh way way slower all right so we’ve got this one done now Let’s go ahead and generate two and instead of G zero we’re going to set that to G1 and that should give us our second card here which is the Tesla P4 and let’s watch that go yeah so we got our little Tesla P4 here And I mean it’s a small card if you look at it it’s got 75 Watts there it’s running at 25-ish Watts right now I mean it’ll get up to maybe 50 60 watts one thing about these they’re passively cooled in testing earlier I did find that by cranking the fans up a hundred Percent I was able to get faster times and Pascal’s pretty old and like these things the link in the description below if you want to buy them from me and I thank you all for any uh purchases you make from my eBay affiliate links because that money comes to me in sports Channel and I love that uh and yeah the Tesla P4 uh probably needs re-padding so I bought some like pad things to redo the the the padding on it we’ll see what kind of difference that makes I hope it makes a good difference but definitely I Think the heat Tolerance on it is not great I mean we’re not really hot right here at 57 yet but I think you’re gonna see it get pretty high I think it’s going to get up there into this high 70s maybe even into the 80s so we’ll keep an eye on that as It’s growing in temps right now I just have the idrac you know running on its own I’m not going to force the fans up to 100 right now because that’s a lot of fan noise and stuff like that I don’t want to do that but yeah that that is a 90 card the difference in Delta that you’re gonna see here is going to be like silly you can already kind of start to see it shaping up and this is the sloppy one here and we got 27 and 60.I mean it’s not what we saw with the other One which was seven and six but it’s still getting there and I mean by any stretch of the imagination this is faster than what you’re plotting with like 99 of the people out there based on the prior plotting competition as a matter of fact I’m like confused at how To do a plotting competition moving forward I don’t know how to do that at this point like if you’ve got ideas about that make sure you uh drop them to me email them to me social digital spaceport.com uh how do you do a plotting competition when everybody is like Literally at the band with the bus of pcie3 I don’t know who wins there I don’t even understand that I don’t think it’s possible for there to be like a this is where you like everybody’s gonna be like I am at the bandwidth with the bus for Pcie3 there are a few people that have true pcie4 systems all the way across like Mike and he’s hitting like 90 seconds so I mean that’s wicked fast times also will there be people that have pcie5 systems like I know those are on the horizon here really shortly the Pcie5 threat I think is what she is seeing they even mentioned this specifically as being a single card pcie5 would theoretically possibly be able to do grind so I mean that’s one of the things I think they’re trying to take on also and prevent from becoming a reality Ron sneakernet for the win hell yeah so biggie this is one of the things that JM talked about is so this hard Fork has a schedule built in just like you know we’ve got kind of the happenings that are pre-scheduled those are predefined so this forecast predefined K uh filter reductions built Into it so they would code that into the logic so it would not need to be in the future worked again for the filter um I agree with you though like I and this is somewhat confusing to me is a case-sized change a hard fork or a soft Fork I don’t know I I think it is a soft Fork but I’m not 100 on that but it would be easier to do that than a hard Fork so I think you know if you’re looking at a hard Fork you’re leaving potentially people behind in a hard Fork What are you gonna name like I brought this joke up it’s too early to joke about it but if there does end up being a hard Fork what are we going to name the the Legacy chain and so I thought xch Prime the one true xch big filters Uh if you if you go back to like Bitcoin days like that’ll make sense to you that’s the name uh yeah uh slowest Time Lord that’s John from cheerlinks excellent resource out there also Farmer’s market.cc if you’re an nft gen this guy is your guy uh yeah they’ve got something called Plow I forgot the name of it but yeah plow is what theirs is called and it will help move your plots to drives really good uh as a matter of fact I need to do a video on the Giga horse or the Mad Max version of that so that should be some Fun awesome two systems you will be using one with a 30 90 and another 38 ETI Bazinga you’re going to be belting plots out just this fast man like it is going to be literally just this fast if you choose k32 size another thing that I did in the Last live stream where we did some plotting and let me make sure this is like showing the terminal right yeah it is okay and me eating pizza great uh yeah so the last live stream we did some k-33s it it’s it’s crazy because usually If you did a k33 a k-34 it wasn’t like in you know the traditional Mad Max platter the CPU one it wasn’t like a linear growth it was like oh it’s gonna like take even longer to get that k33 than double this looks like it’s just double which I think that’s exciting Because that is efficient maybe I don’t know it saves you know some time on the end user to go from k32 to k33 and just be able to say it’s double uh will it be you know double again to get to k34 I hope so I really hope that’s what it Looks like for people out there 50 Watts watching 12 likes smash that like button do it for the outgo bullish Ox you are absolutely right why don’t everybody just uh hit that like button while you’re down there bring that Bell if you are interested hit the join Button join up as a member and you get for 1.99 membership three percent off hard drives that are uh right now in stock on shop.digitalspaceport.com and if you are a 7.99 member you get a five percent discount yeah I think that’s right uh and yeah so I mean those are good deals I try to make it win-win and that makes a lot of sense if you’re buying hard drives because those are expensive puppies uh silicon Lottery basically what do you mean how do you mean that Y2K Anonymous I’m interested what is the context you’re talking in is it like The actual chip on the GPU that is sometimes better than other ones I don’t know okay so we see that phase one here took 425 seconds again like that’s fast that’s like fast fast for somebody’s time uh in normal reality like this is a 90 GPU that’s about to produce a k-32 in The equivalent time of I I don’t even know I don’t know I think you would have to have had maybe 26.96s or something like that v4s uh and blade bit in the past to get into the 11-ish range so I mean this is crazy and this is a DDR3 system DDR3 cheap right Now it’s cheap uh but you know DDR3 right now Dirt Cheap uh 32 gigabyte dimms there are people selling them like on home Lab sales for like ridiculously cheap prices so uh check out Karma and also uh the Mad Dutch dude if you’re on there they both usually have pretty Good supply of them and you can always DM those people and try to find out also at the same time uh yeah if you’ve got a pcie5 motherboard uh then or a pcie4 motherboard your times are going to be cranking out in like the night but here’s the problem Here’s the problem you can you get a desktop system with 256 gigs of RAM the answer is that the desktop systems people powers that be have screwed you and I this is like this so pisses me off no I don’t think you can um maybe there’s some 64 gigabit ddr5 Dimms it’s going to be so freaking expensive man uh so usually you’re looking at how can I get a start and like it’s this daisy chain of stupid because of like manufacturers decisions like oh we’re gonna give you two dim slots or four dim slots like that sucks And that’s where PCI A4 and pcie5 seem to reside on the desktop so hopefully if you’ve got some like threadripper systems if you got threadripper systems you’re like good if you have Epic Systems uh that’s one of the better cheaper ways to get to pcie4 also um I’m unaware currently exactly the Cheap option to get there with Intel because Intel like Skylake and Ice Lake are still like crazy expensive uh if you know let me know xch cash I love it you know I had a lot of bch like yeah I had a lot of VCH back in the day Come on blood come on block come on blood it’s almost done I read our Rippers and there’s aeronixes right there and that man he’s got some systems that are insane his times are good oh you wanna you want the horse race to literally be the silicone lottery for the GPU plotting competition I don’t know I think what might be even more interesting is total like throughput of G like plots per day like how many plots can you produce in an hour something like that seems like a good competition metric also and that’s not like you’re going up against like Milliseconds you know that’s a variable going up against milliseconds I don’t see how that’s a really super awesome competition so okay here we are we are 639 seconds for the Tesla P4 10 minutes I mean 10 and a half minutes like where where was your cpu-based system that was hitting that that was Cost effective or even remotely cheap like to get some of that was pretty expensive with most cpu-based blade bit Ram of course 512 gigabytes of RAM here 256 get you there so that’s pretty cool I mean that is really cool I think that what we’re seeing with the GPU plotting Is not something that scares me I know there’s some people that are like do me about it I’m not do me about this at all I think this is going to be like actually pretty cool and I hope you are having a good time watching along if you are hit that like button Down there get this amped up and also if you want to go and uh make a comment down there if you can I think that also pumps the algo anything to pump the algo just pump the algo 128 gigabytes is hard for a desktop I know thread rippers I mean third rippers Are excellent systems it’s just they’re not they’re coming down in price so I think that they become more cost effective to buy I mean I know some people in our Discord are like watching them daily and sometimes they’ve been posting about them daily also and the prices seem to have come Down a little bit recently um I don’t know if it keeps going down I don’t know if we get the Intel high-end desktop offering anytime soon I mean Intel likes to like hint things and then like maybe they don’t deliver so I mean I’m scared that Intel might even Just be like we’re gonna do this and then like where are they where are the chips are these even in systems that you can ever buy like do they exist like they’ve done that in the past and that’s kind of disappointing when they’ve done that Robbie you just ordered a new GPU for Your gaming rig hoping to put in some used plotting if you ever need to I think you’re gonna I mean I think you’re you’re a GF armor you’re you’re probably gonna need to uh no ddr5 gpus yet so no benefit there’s no ddr5 gpus damn What is going on with the GPU industry I thought the 40 series was supposed to be great maybe the 40 TI series will like fix that I have no idea the other thing that’s interesting is the two-year cycle that Chia chose for kind of the plot filter reduction that they’re modeling I mean What do you think do you think that’s aggressive because I mean it seems to me like the CPU or the GPU release cycle is every three years I could be wrong on that but it seems like it’s every three years and if it is is two years too aggressive I don’t know um Also I think you know the end game that we talked about which is the 400 millisecond potential for lookups on the high side for the 32 resting point for the filter in 2030.I mean that’s a long time off so it’s kind of hard to predict exactly what everything looks like at That point of the game uh that’s very hard to predict but I mean that’s 5.7 lookups 400 milliseconds there’s a huge chance we’re all on ssds at that time or moving to ssds at that time so it could also play out like you know yeah there’s just there’s a lot of Interesting variables at play here there’s a lot of interesting variables and like I said uh if you are interested in that chip if you’re a Chia farmer you should be interested in that chip read it understand it read it three times if you have to like I did understand it and Like literally comment and say what you feel because the chip process is there to solicit feedback from individuals ask questions if you’re confused about things um you know that’s some of the you know kind of guidance that we might be able to get from the people out there and There’s a lot of confusion out there right now about things so try to you know pick your sources of information gingerly and chia officials not bad Max is not bad I try to just pair it information I hear from them as much as possible and not go out on a limb I went On a limb with my math over there but I think my math was right because I I did it a couple times and I think that you know even at resting at the plot filter at 32 400 milliseconds of seek on a disc still gives you plenty of time to be under the Five seconds that you know Chia has targeted and again they need to keep it under that so the chain can keep having that signage Point iteration that’s roughly in the 9.375 second range so the chain progression continues in like a very you know predictable stable fashion There’s no GPU that can come even close to saturate pcie4 are you sure that the oh what is it called the a100 I believe is what it’s called are you sure that sucker can’t come close to saturating pcie because man I think those things are really expensive right maybe I’m Wrong but those are like wickedly expensive gpus and I think they’re kind of freaking high-end I don’t know uh there’s been some stuff said about the a100 I think they have 48 gigabytes of DDR I don’t know what DDR but DDR Ram uh they seem like they’re very fast uh I Don’t know I’ve never used one myself I would love to use one uh let’s see Strider it’s every three years but the release of each model is about two months apart so continual GPU releases okay so now I get it so like do and they usually start out with the high End right like the 30 90 the 40 90 those seem to usually be the ones recently that are like boom they dropped those kind of first on the on the people out there right the 40 series gpus are kind of trash I I have no idea what more I could do with My 3080 TI like I’m like there’s no game I play there’s nothing I do that’s not like absolutely crushed by the 3080 ti so I mean even geoplotting this crashed by the 3080 ti so I mean I don’t know what I would do I mean I would need to Try to invent something to actually like crush with the 3080 TI at this point and I don’t know what that would be um I I would say Mike has got some insanely good times so he’s doing some stuff right he’s got Wild Disco though he’s got like the wildest disco that I’ve ever seen so I mean kudos to Mike if you’re watching still kudos to you for that insane IO that you’ve got it built out over there 24 gigabyte per second Max for pcie4 yeah that’s what I’ve heard is like 12 gigabytes per second Max for pcie E3 24 for pcie E4 5950x 128 gigabytes of ddr4 optane 905p love me obtains love my optane although I’m bummed that Intel is not making optane anymore like that is that’s some that’s something I mean it seems like a really good technology I’m surprised they didn’t like license it out to somebody else or something I mean X-point wildly good stuff 16k gaming yet with the time get with the times uh you know actually I was looking at an 8K camera the other day and I was like holy that could actually push the 3080 TI processing video out of that thing it’s like 120 Megabytes per second uh that it shoots it’s Wild uh is he confusing people current gpus do not use DVR they use gddr yes okay if I said that then I don’t know uh the a100 memory bandwidth is two terabytes per second that’s wicked uh I’ve been looking up on Some stuff on the AMD website and or the uh Nvidia website about them and they look like they come in clusters of cert like on a server for like 16 of them and one cluster Envy linked and I mean Max has said there’s going to be multi-gpu capabilities at some point here for Plotting so it’s like I mean do you need to go faster because I I mean I think Max is in the camp of let’s get you faster um I mean he’s even like themed kind of like that like Mad Max like the Giga horse like his entire motto kind of is Going fast so uh that kind of excites me also about Max I think that he’s wicked cool developer I mean we’re lucky to have somebody of his caliber in the Chia ecosystem as my honest feelings he’s come out with some cool things that have helped out a lot a couple of times I Think this is probably the next incarnation of that looks like we’re about to wrap up that second one here I see lights blinking they gotta be writing we’re about there and what’s up Captain nice to meet you welcome aboard I will get you in the Discord after this so I will uh just Send me your Discord handle and the uh members section and I’ll get you into Discord you’re pretty sure they sold optane to another company or it was just spend off into its own thing I haven’t heard of that I would love that if that happened If you know if you know more details on that let me know Okay cool so there we go 593 seconds under 600 seconds for the P4 to get finished the plot wow that’s insane that is insane right there like do you see what I’m saying when I say The Delta between I don’t even remember 1200 or something stupid like that that I spent on the 3080 TI and a card I spent ninety dollars like that’s that’s insane now we’re gonna run some benchmarks to look at what size Farm in k-32 these might be able to handle What I said diesel am I using so in the 3080 TI over here I’ve got some actually SATA ssds these I’ve been using the entirety of the digital Spaceport Channel show I mean I don’t even know how many petabytes have been written to these but like in the teens of pebabytes Always in a raid always in a gigantic fat raid because I believe in gigantic fat raids for saving your discs uh if they’re ssds or nand uh but I’ve got eight Evo 870 one terabytes if you’re at the 512 you don’t get that generous cache size if you’re at the one terabyte You get a bigger cache and so having those in a RAID 0 is essentially what you’re seeing here for the i o performance on mine I think it’s pretty decent uh if you’re looking at the time there it’s pretty decent um so let’s step over here real quick And uh let’s grab one of these plots there we go we’ll grab this one here and let’s do some checks I’m not sure if the G flag works on the proof of space lookup so we’re gonna find out it either does or it doesn’t um Did not like that proof of space look up G zero okay so I don’t think G is going to be a flag that it’s going to respect I think that’s only for plotting so let’s just see which GPU it picks up here oh I can already tell you which GPU It’s picked up it picked up the 3080 TI so you would be able to effectively Farm significantly more with a 3080 TI well this is interesting actually is it using both kind of looks like it’s using both I don’t know it might be using both at this moment It is sailing along I will say that oh the Intel solder SSD division to solid I’m oh yeah well Intel is data center SSD is locking after a certain period of time and was never my favorite thing it’s pretty fast that was pretty fast for a thousand lookups right there yeah So there we are uh there is a Formula let me go over here and see if we can’t find that formula over on the proof of space for Max’s offering MMX binaries where were you okay there we go and if we go to proof of and so Chia Plot sync is his plot mover that he has the Chia Paws that he’s got here okay yeah it probably did use both of them actually that does look like it actually used both of them so your max Farm size equals your plot size uh times plot filter times eight times One thousand divided by total lookup time seconds okay uh so plus size I guess we would look at a k32 here is that going to be the plot size is does anybody know what the plot size referred to is here because that way I guess we should do a uh LS of this and get its actual size uh 76 gigs okay so we’ve got 76 gigs uh I don’t know even what format he’s using for that is that going to be yeah so I don’t actually have the information I need to actually be able to use this but This is the formula if somebody figures out how to use it that you can use to tell how much you can Farm let me check the chat over here and see if anybody’s actually got that information but this is how you can actually figure Out this is how much I would be able to put onto a farm with this GPU as the farmer at this compression level with this case size level that you’re at Aaron you should be able to find a uh link to the current discount code that is Inside the members area of the YouTube thing so check there I also shoot me an email social digitalspaceport.com and I will help you if you can’t find that out and I will check that after this live stream’s over here and let’s see I think we are almost like There like I don’t know what else is there any other tests anybody wants to see right now I think the Tesla P4 the Pascal series like a 1070 is going to work pretty good in here also I have a really good feeling so I mean the Delta between 38 ETI and the Delta between what we’re looking at with this other they’re so small I mean it’s so small I really really feel like this makes plotting Chia significantly easier for people that are new to Chia and I don’t think it’s a bad thing from that Perspective I know there’s a lot of Legacy users that have plotted space that don’t like this I see a lot of that on Reddit right now A lot of people are like the sky is falling uh I I don’t I don’t get it I don’t see how they’re like Maybe they don’t want to spend you know 90 on a GPU to re-plot their Farm I don’t understand it it doesn’t make sense to me because it seems like to me ninety dollars and whatever time you spend is going to pay you back at this point more than that 90 dollars because Of the percentage increase in your own net space so I mean let’s say you have one petabyte like or one Petty bite you’re gonna go to like what 1.2 pebby bites like that’s a significant amount of daily increased earnings especially at the prices we’re seeing which again at highs for relative recently Uh ten thousand dollars of what you spend I mean I don’t know you guys tell me what you think uh definitely uh make sure that you hit like And subscribe we probably all have some more live streams but nothing else this weekend I’m finishing up some storage stuff on Setting up different types of farm friendly uh rays and also using some software that’s just a little bit more user friendly for folks out there who are not really friendly with Linux and want to get away from Windows because really most people should try to get away from Windows all right everybody Have a great rest of your evening I hope you guys have had fun here today and I will see you guys next time Tesla P4 GPU – https://www.ebay.com/itm/314210522487?mkcid=1&mkrid=711-53200-19255-0&siteid=0&campid=5338899385&toolid=11800&mkevt=1 Join this channel to get access to our Discord, Store discounts + more perks: https://www.youtube.com/channel/UCiaQzXI5528Il6r2NNkrkJA/join Shop our Store (receive 3% or 5% off unlimited items w/channel membership) https://shop.digitalspaceport.com/ Crypto Profitability Tracker Chart https://digitalspaceport.com/gpu-mining-vs-chia-farming-profitability/ 10K 750TiB Chia farm blueprint Article https://digitalspaceport.com/chia-farm-template-750-tib-for-10000/ Chia Earnings Sheet https://docs.google.com/spreadsheets/d/1lXviouJ1dZhT3s_4wQgdySnTMJJX3E5LXcJJv5xftdg/edit#gid=905970511 Be sure to 👉Subscribe👈 for more content like this! Website 🚀 https://digitalspaceport.com Twitter 🐦 https://twitter.com/gospaceport Reddit 🤖 https://www.reddit.com/r/DigitalSpaceport Discord 🕺 (access w/channel membership) https://discord.gg/UEF7fQRDcS Disclaimer: This is not financial advice.Do your own research to make informed decisions about how you mine, farm, invest in and/or trade cryptocurrencies.source.

Leave a Reply

Next Post

Ethereum Price Prediction 2023-2032: Will ETH reach $8000 soon?

Ethereum Price Prediction 2023-2032 [Ethereum Price Prediction 2023 – up to $2,527.58](#Ethereum-Price-Prediction-2023) [Ethereum Price Prediction 2026 – up to $7,956.74](#Ethereum-Price-Prediction-2025) [Ethereum Price Prediction 2029 – up to $25,581.44](#Ethereum-Price-Prediction-2029) [Ethereum Price Prediction 2032 – up to $74,888.90](#Ethereum-Price-Prediction-2032) Cathie Wood declared that Ethereum is more undervalued than BTC and expects an ETH price of US$180,000.This would represent…
Ethereum Price Prediction 2023-2032: Will ETH reach $8000 soon?

Subscribe US Now