Chia mining looks profitable? FORGET IT, if you was ther not in the begining, no way can win it

82

i will say this tho, if somebody wanted to make money on chia, then i think k33 is the way to go…

dug a bit into it, during last quarter last year the chia maintainers upped the plot minimum from k31 to k32 before the release of their mainnet. or whatever…

this was done because they had found out that one could plot a k31 plot in 16 minutes with 6000$ of hardware and thus they argued that it wouldn’t take long before the plotting could be done faster than their minimum of 10 minutes, apparently if one can create a plot in less time than 10 minutes one can cheat the system…

not sure exactly why, but i would assume you can basically plot so fast that you can lie about what solutions you had at the time of them requesting a solution… and thus maybe gpu mine chia… or something like that…

but i digress…

hench they cannot allow for people to plot faster than lets say 15-20 minutes…
since the minimum now is below 60minutes afaik that would mean people could essentially build systems that could plot faster very soon… forcing them to switch to a k33 minimum…

that would invalidate all the k32 plots and thus the ones running k33 plots would get a leg up for a time…
i doubt it will last tho, and i will also not advice anyone to mine chia, i’m not sure what to think about the project … it basically just seems a bit scammy or just a cash grab at best… and if it does work which i doubt, then profits are fully dependent on the network activity.

and sure one can say the same is sort of true for storj, but with storj earnings are from used space, with chia earnings are from user activity or so i would think… ofc they may use like a bitcoin model, but again one is paid in XCH which required market demand to even create value for the coin.

so essentially one can store a ton of data and get worthless coins for it… if one even get any at all…
chia just falls short or doesn’t make sense in almost every aspect i look at or consider…
but still i think the best chance would be k33 plots… ofc if they switch any time soon one might end up with XCH in the wallet… but then i doubt they will have much value because few people will want to replot and thus the chia hype for the next long period will be over…

but its what i would “advise” people to do for chia, if they really feel that its a sensible or profitable project.

plot k33 and wait for k32 to be obsolete, rake in XCH and sell at some future bull market.

83

Relax guys, you were late for the mainnet launch but Chia pools will be available, soon (lol).

84

pools will kill profit at all, as more people will try to mine it.

85

“solo farming” is not profitable atm and using hpool undesirable so it will be a big improvement.

86

Relax guys… This scheme is soon dead :laughing:

Beware of such crypto scams, especially from China!
Chi(n)a coin! :joy:

2 Likes
87

i am sorry you were late for mainnet launch.

88

@SGC where did you read that about cheating the chia system if you can generate plots faster that 10 minutes?

89

I don’t remember where I read it, but I can confirm that I read that as well.

90

I guess the reason might be simple: You have to report a solution within 5 seconds, so if you can generate that part of a plot that contains the solution, within 5 seconds, you don’t need to generate a plot and can compute it on the fly. Could that be right?

91

I don’t believe this is actually the case. You intentionally have a lot more time to submit the final solution so you don’t need beefy machines to farm. They have something called timelords to slow this process down. I think that’s where the 10 minutes come in. If you can tailor make a plot for a specific challenge and you can plot and submit your solution before the time runs out, you can cheat every block. Now you would need massively impressive machines and probably plotting completely in high speed RAM to ever be able to achieve anything close to even an hour to plot end to end. But hardware scales… best be prepared.
Ps. This is based on limited information I have read and I’m not certain about large parts of it. Please do your own further research.

92

5 seconds is the official limit and the software warns you if you need longer as you may lose rewards. Hpool also uses 5 seconds. If more was no problem, they’d probably all do it. Besides, you don’t need a beefy machine for 5 seconds, just idle disks. 5 seconds is a long time for this small amount of data.

93

I’m aware of those log lines, but I don’t think they mean what you say they do. The line clearly says you MAY lose rewards, not that you will.

When timelords are notified about unfinished blocks they get to decide which one to work on based on the number of VDF “iterations” required - the more iterations required the longer it will take. It is possible that a timelord receives notification of a new unfinished block after they have already started the proof of time for a different one - when this happens it will have to look at how many iterations it has already performed to decide whether to continue with the current unfinished block or to drop it and start working on the new one (because it would be quicker).
source: https://www.chiaexplorer.com/what-is-farming

If you’re creating a plot for a specific challenge, perhaps it is also possible to ensure they need much fewer VDF iterations, winning you a lot more time to beat the rest to it.

94

very interesting, I am a little bit surprised that nobody has been pumping VDF difficulty yet. Like with mining, it is fairly simple, suddenly switch on a few EiB, harvest with low difficulty, when difficulty is high, drop out and wait until diff is low again, repeat. Maybe it has not been done because nobody has like 5 EiB capacity to switch on and off at will…

Update: wow, chia network still growing, now at 16 EiB.

1 Like
95

I think storj might pay out this month, chia - not so much :smiley:
Edit : Hooray payout!

2 Likes
96

yes, storj has always been good to loyal SNOs, a bit more egress traffic too for June.

As for chia, it has started a nose dive, we shall see where that is going

97

I am not Fam. with Chia, but maybe plotting local und move to unlimited gdrive is an option. it is only limited to 750gb per day :slightly_smiling_face:
is there any further read/write on one plot?

98

No writes, but they are read constantly. You need hot access to the plots. I don’t think gdrive would work. Storj DCS does, I tested that. But the latency may still be an issue.

99

What does a plot look like? 1 File or 10000 Files?
constantly reading is a problem @ google…
How much plots did you generate until now? maybe its only worth if you do this on a fast dedicated root with 1 or 10 gbit connection?

100

A k32 plot is one big file, about 104GB. The node reads random portions of it.

101

This will result in an API-Ban. Ok. no GDRIVE.
As the Threadstarter mentions there is no chance to win anyway if you start now from scratch. The only chance is or will be a pool…