Gigahorse GPU Plotter, GPU Farmer, Plot Sink: Boost Your Chia Farm by 47.2%! Plot in <2 Minutes

This is rather a big claim, and I would say requires a big explanation.

I am not trying to say that Flex is not suffering (or whatever the reason could be), rather that Flex has few thousand members, and it looks really odd to me that you are the only one seeing such behavior. The math is quite simple, so I would imagine that most Flex members are calculating it every other day or so. I would also imagine that maybe most of Flex (or any other pool) members are using tools like frmr (when it was still alive), chiadog, machinaris, … and some (most) of those tools can provide those stats with a daily update.

(I have nothing to do with Flex, as I moved long time ago to a smaller pool. So, this is not intended to defend Flex, but rather trying to see where you are coming from making such claims.)

I hit 50 blocks in the last 55 days , and 100 blocks in 110 days. This is with 3.95PB. I analyzed the top 10 flex farmers any none of them come close per PB. My node hits 27% more blocks per PB than any of them. I don’t have any other explination for it. It is what it is. I would never switch to a pool when i am statistically hitting 27% more farming solo. Btw when i do chia farm summary it says 1 block every 1 day 4 hours. If you do the math there 110 days * 24 hours per day / one block every 28 hours that is 94 blocks so yea im doing just slightly better than the expected but FAR better than flexpool farmers. Go look for yourself on their miners list.

I hope that some Flex member(s) will go over their data. I still think that it would be easier to just run the data on the whole Flex, as the bigger the sample, the less it is influenced by randomness of those wins. We have here several threads where people were reporting dry spells (I hit 8x or so at that time, some folks had more than that). Another factor is the chia version used, and maybe some of those biggest farmers have rather a hard time following the latest updates (that from time to time break working farms, so it could be understood that it may be better to have a small constant penalty comparing to a catastrophic scenario).

I hear you, but it is really hard for me to imagine that those top 10 members would not be doing those calculations, especially where potential losses are so high.

Again, I have no bone here, so I will pass on it, but would really be looking forward to other Flex members chiming in.

By the way, here is an old article comparing various pools - Chia Pools | One Month Update Based on that, it looks like 10% variation is possible, and this is what I saw last year when I was still with Flex.

I here, which is why i went back and analyzed the last 6 months as opposed to 1 or 2 months because I have seen that short term variabiility myself. But 6 months and across the top 10 should eliminate that. My only thought would be a flaw somehow in the pool in timing for submitting proofs, or just the fact that i dont have those extra steps involved. Beats me.

1 Like

No one would 0.0 Thats crazy good luck.

That being said every pool was measured and compared some time ago and we were all within a few %

1 Like

As I recall those comparisons, in every such comparison, Flex was close to the bottom of the pack. In the link I provided above, it was ~8% behind two best performing pools that month. if I recall dust storm comparisons of various pools, Flex was also at the bottom being more than 10% behind and not showing improvements during next dust storms. So, yes, all that was within a few %.

Currently yes.

Yes you can with Gigahorse.

People go with the big pools because luck variations is a lot less. Big pools are much more stable and give out constant rewards.

People are relatively short-term minded and want the best results NOW!

People don’t want to wait a year to see how variation with Luck works and how everything eventually balances out in the end.

Whether you SOLO farm, or join a big Healthy pool, or join a small Healthy pool, in the end (long term) it all plays out statistically the same in the end.

Flexpool.io is very small compared to the largest pools out there. I would say it would take longer than 6 months to fully compare Flexpool.io with other pools or Solo farming.

One must Solo farm for a year or more and compare your results. Just like getting extremely lucky, you may also get extremely unlucky periods while SOLO farming…

Probably more so, do we hear about those SOLO farmers that do not find a block for months on end.

The 26% better SOLO farmer here is just extremely lucky… in time he/she will find out that it will even out in the end.

Flexpool.io is doing everything right. They are a healthy pool. They just have relatively small capacity compared to the larger pools. They will improve as they get more capacity.

But there is the issue. People are short-term minded (vast majority). So naturally the majority will flock to the big two pools.

I know… it’s sad, but reality…

Months in were slightly ahead of Space but behind some and we were all within 2%ish of each other I believe. Feel free to correct me if I’m wrong as I’m going by pretty old memories. @xchcentral_jm Was it xch central that ran those?

cough cough NOSSD???

1 Like

True, I assumed he was talking about GPU plotting and farming.

Regarding GPU Farming - I have a few GTX1660 6GB Supers and a 64B RAM device. Is there any point in playing around with the GPU plotter or rather not?

Iirc its currently 128gb ram min for gpu plotter for chia chain.
64gb hopefully in the works.

It also says that min 8GB VRAM, but seems there’s a command line argument for less. Just thinking that if it might manage?

Hello, I’m going to buy dell t620 with 2x E5-2690 v2 and rtx 3060ti. How much RAM do I need? 256GB will be enough or better buy 512GB?

Need to wait for 64G RAM mode. Not sure what you mean by RAM device? It needs to be real RAM, not a disk. 6G VRAM is fine with -S 3.

For single GPU 256G is enough, if you want to run two GPU then 512G. You can also run two GPU with 256G (via -r 2) but it’s not as efficient.

1 Like

is the chart on the starter post for plotting only? then how about farming ?does farming c9 plots need a lot more power comparing to c7?

there are a bunch of tables on that here

Question what situation? Max has a working cpu and gpu compression code that works, today right now not just some Vaporware that’s talked about. Until you can run it, for me its just talk.

And where did you come up with this figure? Mr. I just joined 12 hours ago… :joy: :joy: :joy: :joy: :joy: :joy: