Are the only people who can create compressed plots those who have 256 GB of RAM and are currently using Linux with bladebit? Consumer-grade CPUs have a max 128 GB RAM limit. Most people can’t create them because they predominantly use consumer-grade CPUs. Only those with server-grade CPUs and 256 GB of RAM, which typically means only wealthier individuals, can create compressed plots. So, the rich can profit by creating them, but average individuals can’t? Do I understand that correctly? So, the rich have an advantage by creating them long before ordinary people like me can? Are you suggesting that I use GH, which is closed-source, and Max can do whatever he pleases with the software? I think disk plotting only for compressed plots should be prioritized because not everyone has 64, 128 or 256 gb ram.
No point ranting, itll be ready when it is and not before.
You could use GH, but it would seem your against it and thats your choice.
Why would max do anything to hurt his income? Thats nonsensical.
It is not too expensive if you find old workstations that accept server memory
You are complaining about your own choice not to use madmax
And chia is not the one who focuses on making rich people richer…
It is rich people who focus on making rich people richer.
I’m not a rich person, so I decide to use MMX with 64gb ram investment…and now I have 500tib of compressed plots. I grew from 64tib by trading and finding deals that allowed me to expand my farm for zero cost or close to zero sometimes.
Your choices are your own.
I’m not a rich person either, but I have 512GB of ram in a workstation I built it well over a year ago, they are considerably cheaper now. I drive a 14 year old car, the wife’s car is 16 years old, does that strike you as being rich? Perhaps that’s one reason I can afford a bit more for my hardware.
Anyway I replied to your entitled comments on Discord.
i would say GH is for poor/lazy crowd while BB is for more wealthy/enthusiasticish ppl. 
I would say GH was for smart people who wanted to take advantage of compressed plots long before anyone else.
well yeah, but im looking at it from another angle. GH is working with 64/128gb setups under Win10. it has remote recompute feature. it just working with you current setup (mostly), so you dont need to spend (almost).
Chia farming is for “rich” (not-poor) people anyway.
People who are actually poor cannot afford to invest money in a bunch of harddrives hoping they will make money some day, while in the mean time all they do is cost electricity.
400-500$ back in 2021 would get you a system that makes a plot in 30-60 minutes.
400-500$ will make you a plotter today that can do 5 minute plots. When Chia launched building a plotter that can do that would cost many thousands of dollars.
I call that progress for everybody.
But yes, annoying you can’t plot with your (CPU) desktop atm.
HOWEVER, support for less RAM (reportedly 16GB + GPU + nvme) is being worked on, and should be available pretty soon. That will mean that most desktops can then be used, with the only expense (if you don’t have a suitable GPU already)needed a 100-200$ secondhand GPU, that you can sell again after finishing plotting.
I bought a Tesla P4 for plotting (in addition to a 3080 I already had), cost worked out about £90 which is $113, but they are cheaper in the US. In Linux that does 7 minutes plots with 256GB of ram. I wouldn’t recommend one for farming as they seem to have GRRR errors sometimes. You will need a fan to cool the P4 though.
dont fotget lazy part in the statement. i used to use cpu plotter by max since sept 21, it’s easier to follow up same software. moreover i like max’s soft a lot. i’m not happy with fee but i agree that this is chia’s fault, they could be more tolarable and more wise … nevermind
If any of us are rich, it has nothing to do with the money-sink that is crypto ATM. Chia, MMX, others are all some of the quickest ways to disperse a chunk of your disposable income IUAM buying hardware, and so, becoming less rich in the process. HDs are always expensive, electricity is not that cheap anywhere short of DIY and even then, esp. when it’s wasted mostly as heat, and computers still cost $$$.
The difference between Chia, BB vs GH, is that Max produced straight away, mostly workable sw that did what it was billed as being able to do, efficiently, with minimal HW. Chia, Inc. …well they haven’t in many areas done what they said they would be able to do. So one guy who can program has already done it, vs a team of peps that tries to, but can’t quite.
MMX = quick, easy plotting while using most efficient, lower cost hardware, software.
Chia = not so quick, or easy, yet still buggy sw that’s more $$ invested for not even the same result (yet).
If rich peps are smart, they’d be (are already) using GH. They (hopefully) know, that it take money (Max’s fee) to make money. It’s not wasted, but given with purpose. Chia, w/no fee, you pay in a myriad of other ways (lost opportunity, lost time, $$, energy, etc).
Not everyone can afford 1000TB of disk space, should I cry now? Your hole point is nonsense. I bought for 500bucks an old server with 700GB RAM. They are not expensive and if I don’t use it I can sell it.
I get that you’re afraid of Gigahorse, but you can make compressed plots with it on a Raspberry Pi 4B 4GB board. Takes longer, but you don’t have to be a billionaire to get a computer that will make compressed plots. A $150ish Orange Pi 5B can do C5 in under 4 hours, and a NUC at a slightly higher price can do even better.
Check local marketplaces and you can probably get a Xeon workstation that will get you way below 4 hours. I picked up a Dell T7910 with E5-2640v4 (10c20t 2.4-3.4ghz 90W) and 32GB RAM for $100 a couple months ago. It should make a Gigahorse C5 plot in well under an hour.
People who didn’t want to wait six months and then complain… had an advantage without being rich, and were creating compressed plots for half a year already.
Maybe Chia Network should have prioritized something that would take 2-3x as long and have less impact on the ecosystem, but they’ve never claimed to want to be the only creator of Chia-related software (quite the opposite, at least for the last 2 1/2 years–they want to provide reference software and an ecosystem for the blockchain to thrive).
Hey where did you get that great deal???
Me? Local electronics flea market that I’ve been going to, on and off, for ~20 years. But I see almost-as-good deals on local Facebook Marketplace and Craigslist from time to time. I think May or June I picked up a workstation motherboard with an E5v3 Xeon for $150 - it takes 256GB of RAM that I pulled out of a server I got at a junk shop for <$100 a couple years ago.
A lot of the deals I find aren’t easily reproduced, but they don’t even require a comma in your budget, much less a top hat and comically large bag with a dollar sign printed on it.
The one downside to the T7910 is it has the liquid cooled heatsink, so when I add a second CPU the second heatsink will cost more than the CPU itself. 
Most of the hobbyists (small farm) bailed a long time ago when it became clear this was a server / JBOD endeavor. The youtube videos of single PC farming rigs or rows of external drives dried up pretty quickly too once it was shown that scaling out JBODS was the efficient way to organize and power gear (and the $800 coin gravy train dried up).
I personally am grateful the big dogs let me hang here as long as they did, putting up with my silly tinkering and occasional rants. I didn’t farm any Chia at all this year, and have been tinkering with CPU plotting and the MMX Node stuff until finally shutting it all down a couple of weeks ago. I had a blast, but I’m out of my league - I turned my drives into VHDs and mirrored storage space for backups and moved on.
For small farms, passing on compression completely is still an option, as is the MM Gigahorse CPU plotter or NOSSD if resources are lacking - these can all still be done on most office PCs with little stress.
Hey, maybe for old time sake I’ll fire up a node this winter just to keep the office warm and do a little plotting to join you guys again.
It doesn’t have to have liquid cooled heatsinks, they were an optional extra, and worth a lot of money, as is the side cover. The coolers often come up, but the side covers are like gold dust.
I have a pair of E5-2699v3 CPU’s with standard fan coolers, and 512GB of ram, the rear CPU always got a bit toasty when running BB ram plot, about 70c, but I never had any issues.
If you wanted to keep the water cooled CPU, then get a normal one for the front CPU, its only really the rear one that gets hot as the airflow from the front one goes through it.
Thanks for this feedback. I was considering replacing the one with a pair of the typical $50 ones on eBay, and this sounds like it might be the best way to go without getting two new heatsinks.
The “new” T7910 also has the vented side panel; my original T7910 from years ago (dual E5-2650Lv4, 128GB, 4x1tb NVME on Ultra Quad) has the solid panel and standard heatsinks. They’re both going to get upgraded at some point to dual E5-2660v4.
I had to resist upgrading mine any more, its cost a fortune since I’ve had it, but it can generate plots at the rate of one every 2 minutes with two GPU’s.