Ah?? Excuse me. You asked for the debate my friend, yes? That’s why.
But I see now! This isn’t really a debate as you requested, but rather your soapbox to tell us all that you know what’s best. Please proceed - I’m out.
3 Likes
It’s not so much a soapbox as it is questioning. Isn’t that debate? Space for drives is an issue? That is what confuses me. Is that because people here load up server cases with drives? Is it because people put 2 drives in a PC tower and then load up another PC tower? If people say, oh, space is a crucial issue because my server has 12 bays. Well, yeah, duh.
As for “simple maths”. C’mon. Do better. For real. If you pay 2X or 3X the amount I paid to acquire the drive? It will take years and years to make up the difference between me and you before you can say I’m overpaying on power. I would say to you? YOU’RE OVERPAYING FOR STORAGE. (and doing nothing for the environment when it comes to creating more demand for manufacturing more hard drives)
And as for making money? It’s about yield. It’s not about what Chia is worth today. It’s really about the net space and difficulty.
Again this power debate? It matters when Chia is making pennies which is where we are at. When the coin value is meaningful? The power issue is a non issue. Ask GPU miners who use more power in a week than my drives could use in a year. Think about that.
The only reason power is the focal point is because Chia value is pennies. Every cent to power means THE WORLD. Right? Not to me, sorry. You’re talking about using little power, but just by varying degrees. A hard drive does not use a lot of power, by comparison to other POW models. PERIOD. So to me it’s like splitting hairs. A way of justification in my mind.
If you want to talk about heat, go run 2 or 3 GPUs and mine ETH. If you are new to POW, then I suppose you can make it sound like insane heat coming off a bunch of hard drives that are farming. Chia looks up information. Just how heat intensive is this process anyways? I bet few here actually know. After all, you’re just running those big drives so on what basis are the claims being made? Experience? YouTubers? Yes there is heat, but on what scale? Is this like the power debate? Sure we both create heat but it the amount of heat drastic in the first place or is there actually fairly low heat in either scenario? I venture to say heat is not a critical issue because I run GPUs and I can talk about heat. Hard drives? C’mon.
I’m not sitting here suggesting that an enterprise grade farming operation can scale on smaller drives. I didn’t bring up the fact plotting should be done with SAS drives (or via RAM only) because it’s actually doing something good for the environment. But I’m sure people opt for the easy m.2 or SSD option. Faster. More expensive. But again, you aren’t doing the environment any favors. The 300GB SAS drives just go into the landfills. Chia farmers could help reduce that waste, but naw. It doesn’t make sense. It’s better to spend more money and take usable hardware and limit it (take out of circulation) for the purposes of POW.
As for OPEX/CAPEX. You can add in (and I do) the cost of HBAs and cables and be assured that I pay far, far less than you would for any larger capacity hard drive that is being discussed here. I will repeat that operating wise, sure, I’m losing efficiency. But when Chia is at all time lows? Every penny counts. But I’m not farming because of what Chia is worth today. The power consumption is magnified. That’s about it. Power consumption won’t mean S if the price goes back up. If my suggestion didn’t make business sense then I wouldn’t be saying much about it. The CAPEX makes up for the OPEX. But even then, I’m saying you are taking a hard drive destined for a landfill and putting it to use rather than taking perfectly good hardware out of the system for a POW coin. I believe in the spirit of Chia and in that, I know I’m doing the right thing. I stand by the power usage argument that it’s like splitting hairs. Neither method is perfect.
As for the people tapping out, that’s fine. You can’t debate, fair enough. You can’t present examples of space issues? How is it that space is an issue for you? That’s fine. There has been very little to talk about regarding Chia as the price plummeted over the past year. The excitement got sucked out of Chia. So I’m not sure what it is that is considered interesting discussion. Chia needs real life examples of how it is “green” or a better way of POW. There is obviously some real resistance to that.
Start using 100TB disks 
1 Like
Run nine 2TB for two years and you’ve used enough electricity to pay for an 18TB drive.
I used to mine ETH, turned it off recently, so I know what heat is. Yes the server, and 22 drives produce enough heat in the small room without ventilation for it to be an issue in hot weather.
@MisterSavage you really need to open yours eyes and look at the bigger picture, you seem to think that your approach and opinions will fit with everybody else’s.
Mining operations go under when the product price is low. That’s why you don’t want to be running rich but lean.
I pay about ~15% more per TB, and yes, that compensates the cost of a few extra jbods.
I run 9 drives 14TB each. So bigger is better but if a drive fail, you have to replot again. I wouldn’t go under 10TB.
1 Like
I can appreciate that I’m one person on a deserted island. I’m trying to swim upstream, against the current. I fully understand that.
I understand that current wisdom is to go big. It’s logical for the points made here.
I should have asked who is farming in an open air environment. It seems the answer is nobody. If everyone, or most are plugging internal drives into server bays? Of course my suggestion is illogical.
Yes, at some point my power will cross that threshold where your larger drives made up the difference. I haven’t quite decided for Chia, what is the best “look” for them. Is it the idea that drives headed for the landfills are being saved? Or is it better to take perfectly good hard drives out of circulation to farm Chia which will save some energy costs at the Chia end, but what about the cost of manufacturing (energy, environmental costs) to replace those drives take off the market? That is really the big picture question.
I will have some data to bring to this discussion soon if there is anyone left here to discuss things with.
But to clarify a big point here, please do not discuss from the perspective of putting a bunch of 3TB and 4TB into server storage. That is illogical. My whole point here is saving money. That means to NOT spend money on fancy things to hold drives. This is NOT required. Open air is where it’s at. That doesn’t cost money. Yes, a wire rack isn’t free, but in comparison to filling up drive bays on server equipment? This is not a scenario that I’m talking about.
Another consideration is this. Should Chia start climbing again in price, people may start getting into this. With everyone saying 10tb or bigger? It means you’re going to see price increases on those specific drives. Then what? I guess buying now is actually the prudent way to go if that’s your ideal scenario.
I also think it’s worth considering wear and tear. Drive failure. So I legitimately wonder how long these large capacity hard drives will last for farming Chia. It’s like solar panel systems. Do components start to fail and need replacement prior to you getting a return and reaping the benefits of the upfront costs? If you bought large capacity cheap drives, they are used. So a question is whether they were handled well or were run hot by the previous owner.
So to clarify, using smaller capacity drives is not about server cases or fancy things to hold drives. This is about open air farming. It’s about spending the least amount of money on hardware that is possible. The operating costs? There is a negative on that, yes. Unavoidable. But when you compare to other POW systems, the power consumption for hard drives is a drop in the bucket. And no, the current price has no bearing on it. I don’t care what Chia is worth today! I’m yielding XCH. The amount I paid for power to acquire the XCH become less and less significant and the value of XCH rises. At some point it should be absolutely meaningless.
I am efforting to make the smallest possible financial investment into farming Chia. With that, trying to do the best for the environment. I’m going this route because Chia is not on solid footing yet. It’s still a risk. Until they break through? I’m not investing big on hardware. I’ll pay slightly higher operating costs but this affords me the option of shutting off.
I’m not sure who would argue this point. But can we all agree that the #1 cost for farming Chia is the hard drives. A point of agreement? And it’s not close. The accessories/hardware is pale by comparison. The #1 way to save money getting into Chia is on the price of the hard drives. Operating costs? I’ll have more to say on that with hard data.
Lastly, let me try to state this again. If you pay $10K for solar and I use regular power for my POW operation, yes, you could say I’m doing things wrong. However, your upfront costs are MASSIVE. The question because at what point do you pay that $10K system off. It’s an enormous risk if you installed it for your POW operation. I take far less risk. I didn’t put all that money into the solar power. Maybe if the POW operation starts showing long-term promise, more investment might make sense. Solar is obviously a far greater cost than our hard drive situation, but the principle is the same. To me at least. I’ll make a bit less because of operating costs, but damn, I sure don’t have risk. Added to this of course is the risk of part failure on the solar panel system (or hard drive failure in our case) which pushes the ROI out even further.
Legitimate question here.
If large capacity drives make sense, I’m wondering why the enterprise market isn’t using them. You would think that any business that uses servers and storage would want the most cost effective option. I’m curious why enterprise didn’t go big.
If big made sense, and would save power and offer the benefits mentioned here, why then by 2022 has no enterprise moved into that sort of option? It seems odd to me. I thing corporations would push the industry in directions that make sense from a profit/operating perspective.
If it’s cost of production, then are we saying that those enterprise grade 3tb and 4tb drives are vastly superior from a wear and tear perspective from those 10tb, 12tb, 14th, etc. “enterprise” grade Seagate and others? If they call a 14tb drive “enterprise” then I’m curious why these wouldn’t be the go-to solution.
I’m curious on this one. Maybe the answer is obvious, but not for my brain cells. I make no claims of being smart.
What makes you think theyre not?
Which " enterprise " market are you referring to?
Afaik, when larger capacity drives are first released joe public cant even get them, because they all end up in data centres long before being made available to us.
1 Like
Interesting. I don’t claim to know much about the enterprise side of things. I know a few channels that get “outdated” equipment from big companies and this doesn’t result in large capacity drives showing up in retail channels that I’m aware of. I know sas drives hit a wall with capacity. No idea why that is it what exactly makes sas the go to enterprise solution. I believe sas above 5 or 6 tb are like a unicorn. Again I stand to be corrected. I would like to learn as much as possible. The Google’s and Microsoft’s of the world do major upgrades, and often. There are channels that get that equipment and sell it off. I just don’t see big amounts of large enterprise drives. Maybe these are swept up beforehand. I don’t know. Would large enterprise convert to sata drives? I’m sure there are large capacity sas drives that exist but I don’t come across them. If data centers are running 10tb+ size drives? I’m fine with being corrected.
Seagate’s new 60TB SAS drive is the world’s biggest SSD (newatlas.com)
A yottabyte is the largest unit approved as a standard size by the International System of Units (SI) . The yottabyte is about 1 septillion bytes – or, as an integer, 1,000,000,000,000,000,000,000,000 bytes. The storage volume is equivalent to a quadrillion gigabytes (GB) or a million trillion megabytes.
Yes you’re right to try and save money, but if you’re going to be running lots of small drives, then that will easily cost more after a couple of years than buying larger drives.
I’ve just removed two unicorns from my server, a 6TB Enterprise SAS drive (DOM Dec 2018), and a 10TB Enterprise SAS drive (DOM 2017). Large capacity SAS drives are readily available, oh and they run at 12Gb/s, twice the speed of SATA, not that a HDD can utilise that speed. That comes into play when using SAS 12Gb/s SSD’s drives.
WD recently announced 26TB drives, they are currently being trialled with select enterprise clients
From everything ive read its a definite yes that they are using larger drives.
Read some data centre drive reviews maybe, or articles about new large capacity drives, many of which state that they are not available to the general public and are only being supplied to data centres.
It is all about Total Cost to Own (TCO).
- cost of hdds
- cost of equipment to wire/power them up
- costs of electricity to keep them running and cooling
- cost of your time, lost opportunity due to down time, if something breaks/glitches
Everyone’s situation is different. Someone doesn’t need cooling. Someone has cheap or “free” electricity. But generally speaking, energy costs are only going to rise in the foreseeable future, and they will add up.
The economy of scale encourages to use the most efficient equipment. That’s why a 100TB SATA SSD may make sense for some large scale applications, but would look ludicrous for a retail customer.
Thanks for the feedback. I think ultimately it’s up to me to present the setup, the costs, the benefits of saving drives from landfills, etc. Showing the setup and going from there. I do agree that one person’s setup won’t be ideal for the next person. There are different circumstances for everyone.
Regarding high capacity enterprise drives, I think if you go looking for ebay listings, it will be quite clear why these aren’t being gobbled up by farmers. You can just look at 6tb sas and realize the prices are insane. Expensive is never the go-to. Enterprise or otherwise.
I will ask the Chia team at some point what’s more important to them. They ultimately should be advocating for people to plot with 300 or 600gb sas drives. Saving drives from land fills should be something they advocate for. Of course weighing the actual power consumption and which is the lesser of two evils. Considering you can get 300gb 15K sas drives for almost free, this should say something. They need a use and plotting is something they can do rather well. Nobody is promoting that strategy other than Sloth a year ago.
Swimming against the current is never fun. I can vouge for that.
There’s no way I would every get 800 300gb sas drives hooked up to create a pizza oven of heat. Make this more simple for me 54 disks vs 1 18TB. Homie don’t think so. Better yet one
holding 8 18TB drives or 432 300gb disks…My electric bill wouldn’t allow me to buy beer. 
1 Like
If you’re gonna build a 200gb farm with 2gb drives, i wish you lots of luck. End of argument.
I have the feeling that you’re coming at this with a hobbyist mindset. That’s what I can gather from your other posts too. Nothing wrong with that, of course. It’s just different to mine for profit than to mine as a hobby.
1 Like
To clarify, not talking about 2tb drives. To clarify, not coming in with a hobbyist mindset. I’m coming into this with the Chia “green” mindset. For me that has to include trying to save server hard drives from finding landfills. I didn’t come into Chia to piss away money. For some, that ship has sailed. From day 1 it has been about paying the least amount for storage for me. I’m not spending money on things to hold my things that make the actual money. There is plenty of room for innovation when it comes to open air farming.
One aspect I didn’t mention about going with huge capacity hard drives is the risk. Many don’t approach POW like I do and that’s fine. I’m not buying hardware that is specific to the POW unless the coin is proven and is a trustworthy project. Those are far and few between. Chia certainly hasn’t been what I consider solid or a run away winner. So to me, until Chia proves itself, I’m buying the cheapest gear that is the EASIEST to sell. Chia is far from a guarantee. Crypto itself is a gamble. But I would have a far easier “out” than people who invested in server hardware that nobody will want or big hard drives that will have a very limited audience who want to buy them. This is how I make my choices. Other people jump in with both feet and that’s fine. I will scale up into different hardware when I decide Chia has broken through.
At then end of the day, with spending what I spend on drives, my return on investment is far more shorter than if I paid for large capacity hard drives. And that takes into account any cables of HBAs that I required. It was and is for me, the quickest ROI and the least amount of risk.
I still like the comparison to choosing a solar panel system vs. just paying for power. It’s like this debate about big drives vs. small drives and the power draw comparisons. It’s like the debate about buying GPUs that are power efficient vs cheap GPUs. I realize that some people don’t have reasonable power rates so I’m not speaking to anyone in that boat. If you have expensive power, POW isn’t for you. But if you pay $10K for the solar panel system and I don’t, you can claim superiority, always over my choice. Sure, ETH miners who paid 5x the price for an efficient GPU can laugh at me when I paid next to nothing for my ETH mining GPU that is not as efficient. But it comes down to how long do you want to wait for that return on investment? And while you wait for that return on investment a lot can happen. Parts can and do break. Hard drives can and do die. GPU prices can and do change. If it takes you 2, 3, 4,5 years to make up the difference? I say who the F cares. I saved drives from landfills. You reduced slightly the power consumption of your farm. Your choice means drives taken out of circulation for a POW coin. Not much honor in that at the end of the day. A business case for it? Not much of one if you ask me. But that’s me.
You seem very convinced, good for you then.