The root cause of the disease

General Chia Discussion
1

Just like a doctor diagnoses an illness before treatment, understanding Chia’s performance issues is crucial. While discussions on Discord and forums often point to the pre-farm, its impact on price is debatable. In my opinion, compressed plots are the true culprit.
Professional farmers are likely the biggest spot sellers of Chia, with the AMM selling on behalf of CNI possibly contributing as well. Their advantage in cheap electricity allows them to achieve maximum compression and earn over 300% more XCH compared to uncompressed farming.

AA
AA3667×1839 248 KB

The attached graph (XCH.farm | Compressed Netspace Estimate for more information) visually demonstrates the underperformance gap that emerged with compressed plots.

Further to the data, let me share my experience. My initial experiment with compressed plots involved Bladebit C8 (if memory serves). As the halving neared, I replotted with GH C18 and, to a lesser extent, C30. Interestingly, I’m currently earning the same XCH as pre-halving. Meanwhile, professional farmers are likely exceeding their pre-halving earnings.Plot Filter Reduction also seems to have fallen short

So, what’s the solution? Holding back pre-farm sales? While acknowledging its potential price impact, it’s akin to treating a bacterial infection with corticosteroids - temporary relief with a worsening infection. We need to tackle the root cause. CNI should prioritize ending plot compression to get Chia back on track.

Sincerely,
Mojonario

5 Likes
2

Plot compression is not the issue.
An extreme shortage of REAL WORLD demand is the issue.

The solution would be to get serious REAL WORLD USERS interested in Chia, as an example get listed as a payment option on BitPay.

5 Likes
3

We need to differentiate between two cases. Professional farmers, who have immense farms and need to sell a large part of the XCH they generate each month to cover costs such as electricity, salaries, etc., and regular farmers, who mostly hold onto their generated XCH and occasionally sell them when they think there’s an opportunity cost.

Professional farmers compress immediately as soon as the technology is available because it simply improves their ROE. If I have been able to generate the same amount of XCH through compression as I did before the halving… Imagine what professional farmers can achieve with access to cheap hardware, cheap electricity, etc.

This, combined with the fact that compression seems to never reach its limit, is the reason why the total netspace hasn’t decreased as expected with the current price of XCH. Professional farmers now have more costs and also generate more XCH, so they have to sell their XCH on the market… the price drops, then, a high compression level is released, they replot… and the cycle never ends, or at least it seems to have no end.

4

So CNI should stop dumping XCH every month, so professional farmers don’t have so much pressure to dump their coins. Right now CNI is doin’g nothing to improve situation, they are just paying themeselves every month for ‘work’.
Price is going down, cause Chia is useless right now, without clear feature, and a lot of problems that are not addressed. Only thing what is addressed now, is salary for Gene xD

3 Likes
5

Darling there are no professional farmers because it doesn’t make money.

Business example of competent 1PiB Chia farmer. Lets say you are the lucky one, and bought 20TB drives for 18$/TB. Some old SuperMicro server. 20k $. One needs 20+ years of IT knowledge to successfully, and efficiently built this stuff including troubleshooting/debugging.

In last 3 years, I am at about 200hours. If I count the usual IT guy salary of 200$/hour…you are at 40k just in labour. Lets say you are the usual Johnny SixPack working for 25$/h. Just labour 5k$.

BackBlaze guys are confirmation of this example because it is not economically sustainable. PERIOD.

You have now 25k $ investment. HDDs last officially 5 years, BackBlaze stats say HDDs usually last 3-4 years at full load. Chia farming maybe longer, who knows.

You need to earn 333$/month just to pay off HDDs. Nobody has got free energy. Ther usual in Europe nowadays is 0.3$/kWh. 1PiB rig usual burns 500Wh or 360kWh per month or 108$/month.

You need to earn 441$ month/just to pay off HDDs, and electricity.

Add 83$/month to pay of your initial labor, and you are at 524$/month costs. Efficient plotting usually costs around 0.1$/plot. 1PiB is about 10k plots, already 3x replotted. 3000$ for plots.

Another 50$/month. We are at 574$/month just for the basic stuff. With light compression, and efficiency you get around 7 XCH/month after halving + filter halving.

So you are currently losing about 430$ per month. You can dream, like most kids do, about the bright future. Cashflow is king, and only thing that matters.

That is just for the start. You will need once a year maintenance, about 10 hours. If something crashes, as usually does…lets say 50 hours per year.

Add another 1000$ in your labour. In the real world, nobody does business just to survive. Even the 3rd world miners have at least 40% margin. XCH price would need to hover above 150$/XCH to make business world notice.

A large scale farm would need full-time employee. You would need EiB of space to even bother.

your technical analysis is just wishful thinking as every other bagholder in markets. BTC/ETH is manipulated since its regulatory days. Comparing price to other is pointles.

Price reflects costs+utility, and there is none in the whole crypto space. Rest is only paranoia, HFT, inside trading other crap migrated from real markets.

You can think about it as home project to build better financial future because you will most likely never get your money back. Especially, if you wanna do some 30x 2TB farming. Never understood these kids in first place, but boy I have heard interesting bullshit from these kids.

I finally understood why so many like to give me money, they just don’t want money. They just throw it away.

2 Likes
6

Beer for this forum guy! Good points and calculations.

7

Why ASSUME expensive 20TB drives?
14TB are widely available today for $100 well under HALF your figure.
74 drive Supermicro servers can be had fairly regularly for under $1500 WITH shipping (reference MET Servers or The Server Store, though The Server Store is currently out).
You don’t need 20+ years of IT to build them - 5 would be PLENTY, and probably less.
Why are you assuming a crazy labor cost THAT IS NOT BEING PAID TO ANYONE by even a medium sized Chia farmer?

BTW, that’s an “investment” of a little less than $10,000, and ZERO in labor cost when you do it yourself.

$0.3 /KWH is VERY VERY HIGH ENERGY COST. US average has slipped up - to LESS THAN $0.15, and under $0.10 is common with $0.06 achieveable (I’m actually at about 4 cents per KWH but that’s NOT achievable for most).

Electric usage for me on that 74 drive Supermicro (not a hypothetical number when I bring it up, that’s a large part of the core of my actual farm) with the drives is almost 900 watts (measured with a wattmeter, that’s the AT THE WALL real world figure). 21.6 KWH/day is ballpark 85 cents a day or perhaps $26/month.

HDs ROUTINELY will last 10+ years, just having a slow increase of probability of dying over time. But we’ll go with your 5 year figure, so $167 a month to pay for the HDs on average. Add electric, still under $200/month.

You try to factor in a DOES NOT EXIST LABOR COST. Serious STRAWMAN argument, Chia farmers ARE NOT GOING TO HAVE A LABOR COST.
Cost per plot also does not apply, as YOU ARE ALREADY PAYING THE ELECTRIC - so you’re DOUBLE-CHARGING for the plots.

Also, this is going to be more than a 1PB farm - factor in PLOT COMPRESSION - though that will require a pair of 36 drive Supermicros and something like a pair of A2000 (one per) it would also easily give more like 2-3 PB of effective farm size and a slightly higher cost (I’ve also got THIS specific setup, cost about $1100 with shipping).

I’m going to ignore most of the rest, you’ve already demonstrated a crazy number of BAD ASSUMPTIONS that make your “analysis” totally false.

But I WILL point out that, as far as I know, NOBODY has 1EIB of farm.

1 Like
8

This was hard to read. Here is ChatGPT’s summary:

" The argument challenges the assumptions made about the costs and requirements for setting up a Chia farming operation. Key points include:

  1. Drive Costs: 14TB drives are available for around $100, significantly cheaper than assumed 20TB drives.
  2. Server Costs: 74 drive Supermicro servers can be purchased for under $1500, including shipping.
  3. Labor Costs: Assumes minimal labor costs, arguing that building such a system requires only about 5 years of IT experience and no significant labor expenses.
  4. Energy Costs: Claims an average U.S. energy cost of less than $0.15 per KWH, with $0.06 achievable. Provides a real-world example of a 74 drive system using 900 watts, costing about $26/month in electricity.
  5. Drive Longevity: States that hard drives typically last over 10 years but accepts a 5-year lifespan for calculations.
  6. Overall Monthly Cost: Estimates the total monthly cost (including drive depreciation and electricity) to be under $200.
  7. Labor Cost Misinterpretation: Criticizes the inclusion of labor costs as unnecessary and misleading for Chia farming.
  8. Plot Compression: Mentions that plot compression can increase farm efficiency, reducing costs and expanding effective storage capacity.

In conclusion, the argument refutes the initial high-cost assumptions and labor cost inclusion, suggesting a more cost-effective and realistic setup for Chia farming."

1 Like
9

We can’t say for certain that these calculations accurately represent the typical Chia farmer. Their data likely portrays a farmer who succumbed to FOMO (Fear Of Missing Out) early on, purchasing overpriced hardware in 2021 and residing in an area with high electricity costs. Additionally, your data fails to account for the potential earnings of such a farmer if they had implemented compression, which is the very purpose of this post.

When referring to a professional farmer, I envision someone with far greater capacity than 1 PiB. These individuals typically employ high-density storage Jbod enclosures like the Exos E 4U106 or similar. They often operate in regions with significantly lower electricity rates and hardware acquisition costs than one might imagine. Bulk purchases of disks, such as 20 PiB, carry lower per-PiB costs compared to smaller quantities. Moreover, negotiations with suppliers can further reduce expenses.

The intent of this post is not to demonize large or professional farmers; their contributions to the Chia blockchain’s enhanced security are undeniable. I also want to emphasize that this is not a campaign against Chia developers such as Harold, Max or Dr. Nick. My primary objective is to propose the hypothesis that, in my opinion, the emergence of compressed plots has had a far more significant impact on XCH’s price underperformance than other factors. While these other elements, such as pre-farm sales, lack of adoption, and inadequate marketing, may have contributed to the price decline, I believe their influence is overstated in various forums.

10

I think the calculations are spot-on. As you scale up, you see more clearly that this is a losing game.

11

If we assume that farming Chia is not profitable anymore… then how do you explain the netspace?

12

Irrationality. That, plus very cheap electrical regions.

13

People are very slow to react, they have hope, they carry on regardless, they just don’t see what the farm is costing to run. Then there are the people that don’t pay the bill, so they reap the benefits whilst others bear the cost.

I probably should turn my farm off, but my server does other duties as well, I have solar, but this only reduces the cost as I get between 15p and 25p for exports, so electrics still has a cost in lost exports.

I’ve never sold any XCH, so I’ve my farm has always cost me money to run, but it would likely be cheaper to but XCH than farm it, but like most people I’ve got more important things to get on with.

My farm isn’t costing megabucks to run, its a 3060 and 323 TiB raw (less the 5 x 18TB drives which are currently off due to heat issues), so currently I bury my head in a bucket of sand and leave it running.

I am losing interest, I’ve already lost a small fortune - I started in May 2021, and upgraded along the way.

4 Likes
14

That’s a pretty good summary of my points.
I’m also quite sure it would be possible to build a setup that’s even lower cost than mine, though drive costs aren’t going to change and are the majority of cost on most “sane” setups, so probably not a LOT cheaper.

I did forget to mention that the non-drive part of the setup will mostly tend to last a LOT longer than 10 years. and to point out that Backblaze stats tend to not track way past the 5 year point as they’re replacing for higher capacity drives to keep their operating costs down by the time most of their drives get to the 5 years in operation point.

15

darling 1-22TB HDD has always the same wattage 6-7Wh. What sane person uses 100 x 1TB HDD with 700Wh if you can have 5 x 20 TB for 35Wh. You are clearly hobby guy. In business world, it makes no sense, and that is what killed most kids in Chia.

I recall “experts” having AMD EPYC mega plotters with amazing 2min plotting speed…and he used <8TB drives ROFL While preaching about misunderstood DCF from uncle Warren. I am still laughing now. I wonder if this guys still farms.

So you are saying darling that you work for free. Poor people in Swiss work for about 30 CHF/hour. IT guys earn easily 250k CHF/year or 130 CHF7 hour. So you wasted your precious time at your real job…and the think free ROFL

no wonder Earth suffers from raging poverty :wink:

Yes, we have in Europe so high electricity bills. Be happy you pay so little. We are fucked in every hole we have, and paying for it in tax :wink: While sponsoring you guys :stuck_out_tongue:

Yes, I have 20y old HDDs. Will modern HDDs last so long? That is just wishful thinking as DCF. It may last, or it may not. The usual PC stuff is 4-5 years for accounting.

What you mean by NOT EXISTING LABOUR COSTS, please? Do you have some robot that does work for you? If you waste your energy, it is labour costs PERIOD.

Darling plot costs do not happen from thin air. If you are rich, you get < 0.1$/plot because a rig costs money, electricity costs money, and your labour costs money.

It looks like you are women, they do not know the concept of value as well. They believe money grow on trees, and rich are here to sponsor poor.

So the bottom line you are not happy with your bullshit, and you need to explain your illusions. I wonder why? It is your problem. One speaks only when one knows, this is the way darling.

Your “arguments” make no sense, it is just SPS talking. Why do you want us to enlighten with your poverty is unclear.

16

I am aware of large scale guys. The business applies to scale as well.

I have got about 20% discount for small purchases. Large scale guys can get perhaps direct deals. Wholesale prices.

For large scale you also need 5-10k plotter. What you save on HDDs, you “waste” on plotter.

I doubt large scale guys can earn money.

Let’s say you can get 20 TB HDD for 150$. 10 PiB will costs you 512 HDDs * 77k $. It will consume 3.6kWh just for drives. Let’s say you are the lucky one on 0.15$/kWh or about 0.54$/hour

10PiBe earns about 70 XCH/month or about 2$/hour. You are theoretically 1.5$/hour in profit. Or about 1000$/month so just to pay for theoretical low price HDDs, you need 77 months.

That is about 6.5 years just for drives. Where is money for plotter, and your imaginary non-existent labour? Or do you sponsor it from your main job?

AT large scale, you won’t manage alone. You need to pay the guy. No sane person would do this. Usual business strategy in most cases is 5 years. Energy stuff 20y, but they require “insurance” of profit.

17

So you have PV, and most kids believe it is free energy…imagine you just bought it purposely for Chia farming. If I recall, you are from UK and you PV set costs just 10k GBP.

Swiss have more romantic prices…like 30k $ for 17kWh theoretical cap.

PV panels allegedly last 25y. Very few count price of energy. I used to work for hydroenergy guys, milling Pelton & Kaplan turbines.

I highly doubt PV is cheaper than public utilities if you properly count costs of installation, and lifespan.

18

Its far cheaper in the long run, my first lot of PV recouped its costs in just 7 years, its now paying towards the second and third lot of solar, which are also recouping their costs :wink:

Yes I could have something fail, but that’s a risk with anything.

19

Do not forget BB is backup company, their drives run flatout 24/7. Their stats must be “compensated” for your use. The manufacturer 5y warranty is the sweet spot that only most hardcore users can use.

The usual Johnny 6pack will have HDD 20y for his porno needs :wink: I still have some 20y old drives as backup. These drives run perhaps 1 month in 20y.

Anything that spins for 24/7 rarely last too long even with modern tech. Wear & vibration is bitch.

20

I have friend that caught the subsidiaries taxpayers sponsorship. He was repaid in about 5 years if I recall…20y ago gov paid 60% of costs, and utilities licked your ass just to get excess energy.