Debate me. Large capacity drives vs. smaller capacity. Best strategy and why

[seymour.krelborn]

@MisterSavage

The audience here are the original farmers.

How did you determine that?
This web site does not filter on who can see this thread. It is open to all.

@MisterSavage

Nobody likes to admit a mistake

You are conflating what people “like” to do with what people “actually” do.
I do not like going to the doctor. But I do.
I do not like sitting around waiting for my car’s oil change. But I do.

@MisterSavage

Of course everybody is going to defend their decisions.

There is nothing wrong with defending a decision, based on the conditions at the time of their decision.
Defending a decision is not the same as asserting that a wrong decision was a right decision.

I started with Chia when it was at $1,600.
So I defend my decision to purchase hardware to mine for XCH. However, I never asserted that I made the right decision.

@MisterSavage

When I think about it, this discussion was more geared towards people who would be starting out with Chia at the value it is.

100 comments into the thread is when you decided to think about the audience for this thread?

@MisterSavage

This is about building a cheap farming operation that will get you a ROI faster. If the lifespan of the hard drive can’t Outlast the so-called power savings, then it’s just a dumb thing to be discussing.

I own a LaCie 18 TB USB drive.
Let me know how long it will last. Give me the answer as to whether it will outlast its so-called power savings.

Do you have a crystal ball that gives you a list of drive serial numbers to seek out and to avoid?
The lifespan of any drive is a question mark. The same goes any machine.

Less plots = less up-front cost.
Less plots also = less proofs found (less mojo)

If it takes a $1,000 investment 5 years to break even, then it will take a $10,000 investment 5 years to break even, etc. If your farm is 10x the size, then your proofs will happen 10x as often and your mojo wins will be 10x as high.

@MisterSavage

This is about building a cheap farming operation that will get you a ROI faster.

If a cheap farm brings folks a faster RIO than an expensive farm, then those with expensive farms would do better if they split up their expensive farm into multiple cheap farms?

@MisterSavage

Further to that, nobody’s actually come up with any data about actual cost differences or how more outrageous it is on a smaller drive especially when you take into account how much more expensive a new hard drive would be with a warranty or if you roll the dice on a big hard drive with no warranty that’s used.

Most hard drives, if properly maintained (clean power and proper cooling) will last for decades.
If used drives cost ½ the price of new drives, then they would be a bargain.

New drives cost more because they are new, and because they come with a warranty.

Buy a new car, and drive it home. The value of your car just went down 10%. Yet you put only 5 miles on it. It is as good as new, but no no longer as valuable as new. You pay a premium for all things “new”.

The only reasons that I can think of for purchasing new drives are:
– Warranty.
– You know that the drives were not in a toxic environment.
– You do not have to trust that the seller will send you a brick or send you nothing.

Beyond that, the price per TB determines the value of the drive.

Larger drives are more power efficient (it does not cost more to spin a 20 TB drive vs spinning a 10 TB drive). So a cost analysis is needed to see if the higher price of the larger capacity drive is worth the up-front cost (how long will it take for the power savings of the larger drive to pay for its higher cost per TB).

There is a sweet spot, where you can have it both ways – favorable price per TB in a high capacity drive.

You will not have the highest capacity drives with that sweet spot, because you pay a huge premium for the last 2 TB on the highest capacity drives.

Lastly, I suspect that most folks do not have unlimited space for their farms. So they will want to make the most of their limited real-estate by using high capacity drives.

And before you reply, remember:

i am the general i'm runnin' this thing what i say goes

[MisterSavage]

@seymour.krelborn

If a cheap farm brings folks a faster RIO than an expensive farm, then those with expensive farms would do better if they split up their expensive farm into multiple cheap farms?

You don’t know the crypto POW model very well. Not a criticism. Just an observation. You build out. Well, most people do. It’s wise. Let Chia be the shining example. Look at the price! People will die before ROI. Who would have thunk it last year?

ETH mining. If you get a few cheap GPUs to start, and as you near ROI, those GPUs can collectively mine profit towards a new better GPU. And so on. Soon you have 10 GPUs mining that can bring profit towards that new GPU. It grows out. Could you buy 20 GPUs to start? Um, sure?

You build with a very cheap base. If you do it good enough (cheaply) you can eventually justify find more expensive hardware. My main point is nobody around here has a mindset of a $30 chia value. Everybody got into this when Chia was over $100. So you tell me how anyone is supposed to relate. Everybody jumped in with both feet. Made decisions based on a bloated Chia value and great ROI. In hindsight, investing the least amount on hardware and trying to get a quicker ROI makes a lot of sense. To me at least. The proof is in the pudding. The netspace has been going down there is nobody new joining. There’s only one explanation. ROI. ROI. ROI.

But to address your last point about space. I’ll go back to my point. If you can get to an ROI in a reasonable time by being extremely frugal on storage, you can replace older equipment with newer and better equipment. I think that’s called scaling up. Sure, eventually grow into bigger hard drives. The operation can collapse on itself in terms of space. I’m not suggesting a 300 TB farm built with 2TB drives. As a starting point? Do whatever is the cheapest. Not saying I always followed my own advice and I have mistakes in purchases.

[drhicom]

You still haven’t installed that power filter to save you from Con-Ed power sending you spikes every killing your drives

[seymour.krelborn]

@MisterSavage

You build out. Well, most people do. It’s wise.

You are stating the obvious.
I doubt that anyone sizes up their space, and orders the maximum amount of equipment that can fit in their space.

Everyone starts with something, and expands as their budget permits.

@MisterSavage

ETH mining. If you get a few cheap GPUs to start, and as you near ROI, those GPUs can collectively mine profit towards a new better GPU.

That new GPU costs what it costs. Whether you used earnings from the cheap GPU or money from peeling potatoes, you will be spending what you are spending.

@MisterSavage

Soon you have 10 GPUs mining that can bring profit towards that new GPU.

Define “Soon”.

@MisterSavage

You build with a very cheap base. If you do it good enough (cheaply) you can eventually justify find more expensive hardware.

You are under some spell of believing that cash from mining has a different value than case from peeling potatoes. The hardware costs what it costs.

When you pay for your new GPU, the store gets paid, and they do not care from where you derived the funds.

If you purchased better hardware from day 1, then that would have a better earnings from day 1. It also costs more from day 1.

When you purchase less, you earn less.
When you purchase more, you earn more.

Whatever hardware runs the most efficiently will give you the earliest return on inventment.

@MisterSavage

My main point is nobody around here has a mindset of a $30 chia value.

Please put a link to the poll that confirms your assertion.
Or are you speaking for everybody?

@MisterSavage

Everybody got into this when Chia was over $100.

Please provide a list of everybody, so that we can confirm that your conclusion is correct. You seem to have information on “everybody” that no one else has.

@MisterSavage

Everybody jumped in with both feet.

More of the same “I know about everybody” rhetoric.

@MisterSavage

In hindsight, investing the least amount on hardware and trying to get a quicker ROI makes a lot of sense.

You wrote it on the internet, so it is true.

@MisterSavage

The proof is in the pudding. The netspace has been going down there is nobody new joining. There’s only one explanation. ROI. ROI. ROI.

Point me to your proof.

@MisterSavage

If you can get to an ROI in a reasonable time by being extremely frugal on storage, you can replace older equipment with newer and better equipment.

Define “reasonable time”.

Also, see above on hardware costs what it costs.

@MisterSavage

As a starting point? Do whatever is the cheapest.

Buying cheap equipment is the most efficient way to maximize earnings?

Those that have expensive equipment should replace it with cheap equipment? Right?
Yes, they already have expensive equipment. But you are pushing cheap equipment as the best value. So lots of cheap equipment is better than lower quantities of expensive equipment?

You are all over the map, shooting from the hip, making it up as you go along.

i find your arguments strewn with gaping defects in logic

[MisterSavage]

Beam me up Scotty, there’s no intelligent life down here.

(okay, that’s me kidding around)

You’ve humored me so I will counter some of your well thought out, well presented thoughts/arguments. Like a child’s bubbles, I look forward to popping these one at a time.

[BadgerStork]

Let’s try an example…
A 1tb drive can be found for $5 if you are lucky and don’t mind auctions.
To plot 1TB adds on 50C to $1.
Electric costs are $6 per year based on 3W at idle.
Earings from 1TB are at 40C per TB per month, so in this example you lose money on every 1TB drive.

You can either find cheaper electric or switch to 2TB drive. Getting more 1TB will not help!

[DanRelfe]

To be fair, 1TB drives is probably an extreme example, but…

Just did some quick and dirty, but accurate calculations.

Assumptions:
Domestic kWh rate used is $0.12 (probably fairly reasonable but I don’t reside in USA so not sure)
XCH price as it is now, $42 per XCH
XCH price and Electricity price stay the same over the 5 year illustration
Power consumed per drive based on largly idle useage, spec sheets and from my own measurements
Disk prices from diskprices.com, assuming purchased new, cost of drive covered in Year 1
Plots per disk based on calculations (a few might be off slightly as I don’t personally own all sizes)
Income of XCH taken from chiacalculator.com

So there we have it.

As can be seen, 1TB & 2TB drives make a loss in the 5 years.
18TB drives make the most profit.
20TB drives not cost effective yet due to price premium, but still more profitable than anything less than 12TB.
6TB drives represent decent mid-range value, being slightly more profitable than 8TB drives and earning 42% of the profit of an 18TB drive for only 33% of the capacity.

Profit in descending order are:

You can read any and many things into the above but these are sound calculations as I see them.

Bottom line If you have to pay the above prices for disks and the above costs for electricty, 18TB is the best bet. But not everyone does, YMMV.

[BadgerStork]

Unfortunately, you forgot the part where mistersavage advocates CHEAP TBs.
You have used new prices

[DanRelfe]

I know, I stated above all the assumptions. Used prices in my experience don’t offer such a huge discount as you might expect, particularly when there are deals on new or when you buy in bulk.

My general view is that at best a used drive is about 75% the cost of a new one, certainly if the used market in the UK over the past year is anything to go by.

If we can get concensus on what fraction of a new price a used price is, I’ll plug those figures in and post the results.

[DanRelfe]

@DanRelfe

If we can get concensus on what fraction of a new price a used price is, I’ll plug those figures in and post the results.

Ok I went and big fat did it anyway. The following looks at the profits over 5 years if the purchase price of the drive is 33%, 50%, 66% or 75% of the new purchase price (as per previous table, also included in the right hand column for reference). All other assumptions are the same, but of course there’s no gaurantee that a used drive will last the 5 years (there’s no gaurantee a new drive will last 5 years either, but it is more likely than a used one…);

[DanRelfe]

Just for the hell of it… (I have periods of boredome this morning…);

You can see that 6TB ROI the fastest, followed by 4TB and then 16TB, 8TB etc.

[Ronski]

Nice work @DanRelfe but change the electricity price to £0.29 for the UK and see how that alters things please.

[DanRelfe]

Here it is, based on £0.29 ($0.35 at today’s exchange rate)

5-Year Profit;

And ROI;

You can see the negative figures for ROI years - essentially they will never ROI. This includes 1, and practically speaking 2TB drives. It would actually be worse as UK prices I find do not correlate directly to the USD equivalent - they are typically even more expensive in the UK (unless bought in bulk of course…).

6TB drives again come out on top, with 12-18TB close behind.

[MisterSavage]

Thank you for the extreme break down. The numbers, or the start of the argument are based on fallacy. Well, again I will state the obvious. I’m not talking about 1tb or 2tb hard drives. So if you want to base an argument on bloated small drive prices? Then it makes it easy to sound like it’s a money losing proposition. Hint: you are doing things wrong if you are paying the prices in your example. I wouldn’t advocate that in the least. Try harder. Find better pricing. If you can’t get drives at least 2x cheaper than new or bigger drives? Holy crap. What an asinine assumption. No offense. I don’t even look at new prices. In the used market, if I compare, then it’s not hard to get 2x or 3x cheaper. So any half-wit could get 10tb of storage for the price of 20tb storage. But in my case, I can get 30tb of storage for the price of 10tb storage. On that scale. It’s not impossible at all! Redo you misleading charts. Great charts but the cost of the drives is simply foolish and no basis on reality.

This goes back the the questioning about scaling up or hardware costing what it costs. The fact is this. It takes time and effort to source out hard drives. Of course none of what I’m spoken about here from post 1 is done unless you are getting dirt cheap prices on hard drives.

So to put this in plain terms, I’m getting hard drives in the magnitude of about twice, or sometimes 3x cheap than bigger capacity drives. If you are talking 10tb or more? Then I might be getting 4x. So in other words, in a worst case scenario I can get double or triple the storage capacity for the same price, and YES THIS INCLUDES CABLES AND HBAs.

I’m not sure the power draw numbers. Based on what? A 50% load on a Platinum or better PSU? A server setup?

The proof that nobody new is joining Chia? Netspace is the exact measurement to tell you what is happening. How do I know that nobody new is joining Chia and how do I make assumptions about people around here? It’s because I’m observant. I can read. I can see who the regulars are. I’ve read information online. I’ve tried to find videos. EVERYTHING IS OUTDATED AND STALE. The netspace confirms the fact that new entrants are like unicorns. For now at least. Not that I’m really complaining. Yeah, it’s obvious most everyone here jumped into this with a $100+ Chia. If you want to dispute that then I guess common sense is in short supply around here. Have you heard of something called indicator. Indicators? You make it sound like I’m pulling observations out of my A.

As for hardware (GPUs) and building out, you don’t know what you’re talking about. Cheap GPUs don’t grow on trees. You don’t just find 10 cheap GPUs. You find them slowly if at all. During the real ETH madness you really needed to fluke out. You could be a dummy yes, and buy retail like many Chia farmers did here. Then realize it might have been a pretty dumb investment. It’s not hard to figure out when you read all the bitterness online (not lately) that most people just went out and bought new retail drives on the basis that Chia price was solid at $200, $400, $100, etc. Smart would have been what? Probably slowing down and ensuring as little expense as possible. Why so much bitching online about Chia? Because so many people sunk so much $$$ into it and those people say price tumble and looked at their ROI and ULTIMATELY SAID F THIS S.

If you cannot figure out the obvious then not much sense talking about anything. Jesus H C. If you want to debate when people (99.9%?) of them here started farming Chia at a $100+ price? Seriously? Can you not read charts and understand netspace and what that means? Do you see beginners around the forums and online? You see video comments that were from recent times? Get a grip. ETH GPU miners laugh at Chia farmers because the ROI is a joke. If you approach Chia farming like it’s at $100 like most did here, then you make unwise decisions.

[DanRelfe]

@MisterSavage

Thank you for the extreme break down. The numbers, or the start of the argument are based on fallacy. Well, again I will state the obvious. I’m not talking about 1tb or 2tb hard drives. So if you want to base an argument on bloated small drive prices? Then it makes it easy to sound like it’s a money losing proposition. Hint: you are doing things wrong if you are paying the prices in your example. I wouldn’t advocate that in the least. Try harder. Find better pricing. If you can’t get drives at least 2x cheaper than new or bigger drives? Holy crap. What an asinine assumption. No offense. I don’t even look at new prices. In the used market, if I compare, then it’s not hard to get 2x or 3x cheaper. So any half-wit could get 10tb of storage for the price of 20tb storage. But in my case, I can get 30tb of storage for the price of 10tb storage. On that scale. It’s not impossible at all! Redo you misleading charts. Great charts but the cost of the drives is simply foolish and no basis on reality.

This goes back the the questioning about scaling up or hardware costing what it costs. The fact is this. It takes time and effort to source out hard drives. Of course none of what I’m spoken about here from post 1 is done unless you are getting dirt cheap prices on hard drives.

So to put this in plain terms, I’m getting hard drives in the magnitude of about twice, or sometimes 3x cheap than bigger capacity drives. If you are talking 10tb or more? Then I might be getting 4x. So in other words, in a worst case scenario I can get double or triple the storage capacity for the same price, and YES THIS INCLUDES CABLES AND HBAs.

I’m not sure the power draw numbers. Based on what? A 50% load on a Platinum or better PSU? A server setup?

The proof that nobody new is joining Chia? Netspace is the exact measurement to tell you what is happening. How do I know that nobody new is joining Chia and how do I make assumptions about people around here? It’s because I’m observant. I can read. I can see who the regulars are. I’ve read information online. I’ve tried to find videos. EVERYTHING IS OUTDATE AND STALE. The netspace confirms the fact that new entrants are like unicorns. For now at least. Not that I’m really complaining. Yeah, it’s obvious most everyone here jumped into this with a $100+ Chia. If you want to dispute that then I guess common sense is in short supply around here.

As for hardware (GPUs) and building out, you don’t know what you’re talking about. Cheap GPUs don’t grow on trees. You don’t just find 10 cheap GPUs. You find them slowly if at all. During the real ETH madness you really needed to fluke out. You could be a dummy yes, and buy retail like many Chia farmers did here. Then realize it might have been a pretty dumb investment. It’s not hard to figure out when you read all the bitterness online (not lately) that most people just went out and bought new retail drives on the basis that Chia price was solid at $200, $400, $100, etc. Smart would have been what? Probably slowing down and ensuring as little expense as possible. Why so much bitching online about Chia? Because so many people sunk so much $$$ into it and those people say price tumble and looked at their ROI and ULTIMATELY SAID F THIS S.

If you cannot figure out the obvious then not much sense talking about anything. Jesus H C. If you want to debate when people (99.9%?) of them here started farming Chia at a $100+ price? Seriously? Can you not read charts and understand netspace and what that means? Do you see beginners around the forums and online? You see video comments that were from recent times? Get a grip. ETH GPU miners laugh at Chia farmers because the ROI is a joke. If you approach Chia farming like it’s at $100 like most did here, then you make unwise decisions.

What is it, exactly, that you don’t get ? It’s evident that you have not grasped a single thing I stated above…

[MisterSavage]

My point about your data is simple. Have you spend any time actually seeking out deals on anything? If those are the best prices you can find then I’m afraid you’re doing something wrong. There is a common sense aspect to this. You need to be getting 2x the capacity (at least) for the same price to make this work. Otherwise why would I advocate buying small drives? Why would I buy smaller drives if it wasn’t at least DOUBLE the capacity (yes, including cables, HBAs)?

You tell me. Did you spend time seeking out used drives? I doubt it. I don’t expect most people to put in the time that I have. I can appreciate that most people don’t like to get their hands dirty. Obviously every market is different but eBay has many options as well. I’m not going to spell out where or how to get 2x, 3x or 4x the storage space for the price most everyone else is paying.

Sorry, also, a big portion of my post was replying to that guy who wanted to quote every sentence that I wrote.

Don’t get me wrong. You brought in real data. But unfortunately it’s skewed. Go ahead and redo it with a price for used that is 2x cheaper than used. I’m not interested in discussing new drives because it goes without saying that it’s not the way to make your money back. At least not in ones lifetime. I respect that you decided to post something useful. I don’t think I’m being unrealistic in saying that 2x the capacity should be the starting point. If that wasn’t the case, I’m not sure I would have anywhere near the same conviction on this subject.

[DanRelfe]

Ok, I give up too.

Re-read and re-read again the above, and if you still don’t understand the figures then that’s your issue.

[Ronski]

Thanks. I have 300TiB, most is on 16 and 18TB drives, one or two 14TB drives left

[MisterSavage]

Right. 6TB offer the best profit. Are you upset because I’ve criticized the most key ingredient? The price of the drives? You used a website to garner the cost of the drives and then created an argument based on those prices. If the prices are WAY off from reality, then the rest of the data is meaningless. Sorry. The conclusion is based on fiction. Look at the price of 2tb or 3tb drives in your chart. You saying I’m reading that wrong? I don’t see $10/TB or am I wrong? I get that different regions have different prices. This was the easiest place for you to get drive prices. Ok. But it’s useful in what way? You’ve come up with a conclusion that really has no merit. I’m sure you could “replot” the chart with numbers you find on your own rather than a website summary of prices. To get good prices on hardware it take some doing. It’s not for everyone. GPUs during the ETH mining boon for a good price? Didn’t really exist but you could sometimes luck out. Same thing with hard drives. Many people want to draw conclusions based on their own lack of motivation or hussle in order to get the deals on hardware. Like the charts above. Numbers based on lazy. Numbers not based on personal experience or findings. Sorry, random websites telling me the price of drives? Talk about lazy. And misleading.

[Ronski]

What size farm do you have? What is your average price per TB? How much should we be paying per TB?

I don’t expect for one moment you will answer that.

You think you are so clued up, make out you are so experienced, yet you couldn’t even figure out how to plot from the CLI, I had to explain it.

The title of this thread is a joke, you never intended on a debate, so many people have tried to debate, yet it was always going to be your view that was correct, and everyone else is wrong.