Bitcoin Forum

Hey guys,

I've requested the quote for the mining rig me and a friend are putting together and I'm real stoked.

I did however come across a post in the economy section that called us miners rubes for mining instead of just buying.

I don't have the exact figures but given the increases in difficulty and the decreases in bitcoins mined, even with the increasing price, isn't it better for us to just buy and hold than to build mining machines?

We're laying out about $2600 for the rig, its pretty cool imho and there are other benefits, but I keep thinking there might be a smarter way. Among the other benefits are that I don't have to go through the capital control / tax issues of my country if I mine.

What have been your experiences of this if you are a long term miner? (buy vs mine (?) )

long haul vs. short term
mining is a long term investment it may take 2-3 months to pay off the rig but once you do its mostly profit
if you by then you have to sell or hope it doesnt tank be for you do
with mining your rig just takes longer to pay off
as long as your making for then it cost to run the rig your in the black

First off, I'm not a long-term miner.

There's a rather simple answer though... It's fun, interesting and a learning experience, to put together a big mining rig. The custom casing, airflow modding, fine-tuning GPU clocks and tweaking software.

I don't know if a $2600 mining rig will earn itself back for you. I also don't know if you'll make a profit on buying $2600 dollars worth of Bitcoin.

If only one could peak into the future  ::)

No, but hopefully aware that mining will return most probably less money than buying BTC directly, should the BTC <--> USD ratio go up further.

BUT:

Should the BTC <--> USD rate go down, you still have some value in hardware, you can use your hardware for something else and also it might be a free investment if you have your GPU in your PC anyways already.
Buying new mining hardware just for mining is however a, let call it "very optimistic" investment if you want to invest in BTC and could buy plain BTC instead too.

Wow, $2600 for a rig?  You do realize that just to break even, with no difficulty increases, it would take months?  And that if difficulty keeps going up at half the rate it has, you will easily be at least $1000 and more likely $2000 in the hole?

Mining is a lower risk investment than speculation because you retain value in the equipment. Because of this, investors in mining are naturally going to see a lower rate of nominal return because they are taking on less risk.

In standard investment theory return on investment is a compensation for risk that cannot be diversified away.

The theory of "just buy coins" continues to crack me up. Where will they come from? I'm sure there's a dedicated group of miners who would just love people to "just buy coins from them" as you don't mine and go to work for cash to give to them.

Freakin-a man. Get to mining. If it doesn't work you have some gear and probably had fun and learned a lot.

As long as difficulty is above 1, this means a new block is being generated ~every 10 minutes (for this you need far less than a singe GPU miner worldwide).
If there are more miners, the coins will not be generated faster but only in a more secure way, that's all. I doubt that ensuring the supply of new coins should be a reason for anyone to mine.

Yes.

See my spreadsheet, Invest or Mine? (link in sig).  That'll tell you whether it's a better choice to mine or buy, once you input all of the information properly.

I will buy the "fun" angle, but the learning part seems like a stretch.  What do people learn from mining?

How to configure poclbm?  Wow, they learned to install a simple program and run it.
How to install latest ATI drivers? Woohoo!
How to stick PCIe cards in a box and connect a power supply?  Most computer-inclined kids learn to do this before they are eligible to drive.
How to look at a list of hardware, choose what to buy, and whip out the credit card and order it? Sorry, being a consumer isn't a unique skill.

Other than the few folks here who actually wrote some OpenCL C, verilog, or similar such work, I don't think there is much learning that comes from mining.

Put together a simple model with increasing difficulty, or use SgtSpike's pre-built spreadsheet. How long until you plan to break even?

Rising difficulty is really, really going to diminish your profits in short order. The only way around this is if the exchange rate goes up. Once you're banking on the exchange rate going up significantly, it might not make as much sense to invest in mining hardware rather than BTC.

For the record, I don't expect the exchange rate to keep going up.

God do we really need to have another one of these discussions?

They're both fine options, and every argument either way presumes some sort of knowledge of the future that the arguer does not actually have.

If you expect bitcoin to blow up exponentially in the next few days/weeks then obviously buy coins, if you expect bitcoin to remain stable(ish) then buy mining stuff. If you expect bitcoin to die a horrendous death, maybe you shouldn't be here.

Buying coins is easier, building mining rigs is more interesting for enthusiast/hobbyists. Find your passion.


EDIT: Thanks JJG for demonstrating my point. For the record you've been saying the same thing for a couple months now, and while you might be right in the future, you've been wrong up til now.

One interesting thing to note, last difficulty spike was painful, very very painful. And as a natural result, it has discouraged a lot of people. Last difficulty bump ~70%, next difficulty bump, ~17%. Things are not arbitrary.

Someone needs to sig that. 

Tell me one point in time where it made more sense to buy a 2600USD mining rig (which has costs to operate too!) than to buy 2600 USD worth of bitcoins!

What do you mean by that exactly? Are you saying it doesn't matter how many miners or gpu's you have cranking?

Well said.

It won't make the network generate fewer bitcoins if there are fewer miners. It won't make the network generate more bitcoins if there are more miners. (both statements within some limits, as difficulty is increased/decreased only every 2016 blocks)

Your personal outcome is determined by the ratio between your hashing rate vs. the global hashing rate, so it matters how much hashing rate you have in the sense of how many of the globally available bitcoins you'll get. It won't affect the generation rate though, so you can not (or nearly not) influence how fast (or if) bitcoins are being generated, as a single GPU is enough to keep the system running in stable minimal mode.

The rate of block generation across the network as a whole will remain [very] roughly constant.

If half of the miners turned their hardware off tomorrow, then a few weeks from now the difficulty will drop and block generation will remain the same.

When you turn your miner on, you're not generating additional BTC on top of what already comes out of the network (technically you are until the next difficulty update, but that's beside the point), you're just claiming a fraction of the BTC that is going to be generated no matter what.

The more miners rush in, the smaller your fraction gets, but the same number (again, very roughly) of BTC will be generated.

So what if that one lone miner decides to keep his 50 BTC for himself instead of selling them? Well I guess that would lead to withdrawal of people selling/buying bitcoins but then there will always be some poor idiot to buy them finally once the value surges enough. I wish him luck trying to profit on it the same way the lone miner did, eheh.

Actually I'm really curious about the answer, might be that at the time where BTC were worth nearly 10 USD 1 or 2 difficulty steps back it might have been a good choice to do mining. I'm not sure if it would have been a good choice to set up a minig rig though, as the exchange rate went down after that again and difficulty went up.

Buying bitcoin now will yield only the increase in price at whatever future point they are spent. Buying a miner now will yield the miner, the cost of the miner in bitcoin at some future point and whatever bitcoins can be generated cheaper than power consumption for the lifespan of the miner.

Also mining is a far more resilient means of investing in BTC since mining share goes up when price falls. Remember the war for a widely used alternative currency is far from won, and probably will have some bumps along the way.

Define the parameters of your purely hypothetical situation. When I buy mining stuff in this scenario what is my outlook? Am I forced to play by the rules you coin buyers typically set for someone buying mining stuff such that they must sell coins the instant they are mined?

Actually I will set the parameters myself, if you don't like them, too bad for you.

Theoretical mining rig(s): Caseless -- Used 5870 $175 each // 400MHash/sec; Random mobo choice -- MSI 790X-G45 AM3 $90 each new (cheaper used) // 5PCI-E slots; Cheapo 1x PCIE extenders $5 each; Corsair tx950 $150; some crap ram $20; Sempron 140 $40. There is a thread floating around somewhere about how to build a custom wood enclosure for like $30 of wood and nails that houses 5 x 2 cards very nicely.

2x above (5 5870s) = $2400. Use the last $150-$200 for some fans or something, I don't care. Hashing power of system 4GHash/sec, electricity cost estimated @ 1296kWh / month + fans, let's round it out to 1400kWh/month, locale dependent, 11cents / kWh = $154 / month.

So, I can choose any point in time? Let's go with 2/18 - 4/18. Nice 2 month period of time. This is winter time so the cost of cooling will be opening a window and pointing a fan out of it to vent the hot air.

Miner Start: 0 BTC $2600 Debt $2000 Assets; Buyer Start: 2600BTC $2600 Debt $0 Assets
Difficulty 2/16 ~45000, Miner Generates 1300 BTC in 2 weeks.
Difficulty 3/01 ~65000, Miner Generates 900 BTC in 2 weeks.
Difficulty 3/14 ~80000, Miner Generates 700 BTC in 2 weeks.
Difficulty 4/01 ~80000, Miner generates 700 BTC in 2 weeks.

End of 2 month period:

Miner End: 3600 BTC $2900 Debt ~$2000 Assets; Buyer End: 2600BTC $2600 Debt $0 Assets

Bitcoin Value Start $1 / BTC; Bitcoin Value End: $1 / BTC

If they both sold at 4/18:
Miner Net profit $700 (excl assets), $2700 (incl assets). Buyer Net profit $0.

Source: MtGox all time value list. Bitcoin.sipa.be network computation graph 8/10 - 6/11, bitcoin mining calculator.

If you continue the trend then the miner continues making more and more coins and as bitcoin value explodes pulls even farther ahead. Because I'm lazy I'm going to extrapolate from 4/16 to 6/1 using the 6/1 difficulty (incredibly unfair to the miner) of 434000 for for the remaining 6 weeks. @ 434000 difficulty miner generates 67.5BTC / wk, or 405 more BTC.

on 6/1
Miner: 4005 BTC $3125 Debt ~2000 Assets; Buyer End: 2600BTC $2600 Debt $0 Assets
Bitcoin Value Start $1 / BTC; Bitcoin Value End: $8.7 / BTC

Miner Net: $31718.5 (excl assets); Buyer Net: $20020

Any time when bitcoins are profitable but not exploding in value mining is a superior cost-benefit.
Again, I'm not arguing that either one is superior, just that I'm sick of fanboys touting one method or the other. No one knows the future. Stop pretending you do. If you have chosen one or the other that's fine.

Thanks for calculating and writing out what I was too lazy to do bcpokey.  ;)

I've been meaning to write something like that out for a while, as I keep seeing this argument pop up over and over, finally had some free time to sit down and do it heh. I will mention that I may have written that out a bit more grumpily than I should have as I'm sitting in my room that's like 80F from all these stupid mining machines  :D

As implied earlier, don't underestimate liquidation value of parts. I bought 5850s @ ~$170 new and had no problem selling one @ ~$150 in BTC last week (the BTC of which has appreciated well beyond the $170 I paid). 4 garbage PSUs, CPUs, MoBos & RAM/HDDs I had lying around doing nothing came to about $700 (note I bought one excellent PSU, MoBo, HDD, etc. to be my own personal gaming rig whenever I felt like it) to run 7 5850s which can be sold @ $1050. Total investment of ~$2000 (-$170 because I sold one card) after I purchased a dedicated "real" fan for each setup means I only needed ~$400-500 to get into the black which took about a week.

Thanks for this but it's still stupid to start mining now rather than investing. I experienced this first hand and many know this already. I've put the money where my mouth is and bought more btc with my income along side my old miners and not add to it. If you already are mining, buying is a great way to spread out your risk as well and doing both isn't a terrible idea. Personally I wish I invested instead the first time around building a miner, but it's too late for that now. I can only correct the mistakes of my own past from now on.

I mean it's so much easier to just buy and hold as well rather than suffer through all the heat and noise along with keeping them up and monitoring time to time. If I had just thought it through before jizzing my pants and buying miner, I would have been about 30k~ more richer now after the growth in the past 4 months.  Lesson learned the hard way.

To the 400 MH/s starting from 2/18:
No way a 5870 would have gotten that hash rate back then, as BFI_INT was not in miners yet (as well as the phatk kernel)! Also you did not calculate potential downtime (since you slightly undersized the PSU for example at your projected electricity use of nearly 2000W per hour) + setup time. Also prices of 5870s at that point of time were likely different from 175USD. Also not included is shipping time/cost plus risk of hardware bein DoA (used!) and setup time.

However thanks for that detailed overview anyways.

This was a wholly theoretical exercise in mining vs buying as a response to someone. I chose numbers to most benefit my argument, and I said so from the start. However it was pretty close so as to be relevant, there is no point in taking shipping / building time into account, as the idea was to begin mining on the same date as buying coins, not start buying equipment. The chances of DoA equipment is irrelevant. DoA rates are negligble. I've been continually buying 5870s for < $180 for the past month, there was a sale from newegg about 2 or 3 months ago for 5870 2GB models for $180. It's not at all that much of a stretch. A 5870 gets between 430 - 440MHash/sec right now with BFI_INT and phatk, 400Mhash with a good overclock is reasonable. 2000Watts (per hour doesnt mean anything) is overkill on electricty. I run 3x5870s on < 600Watts right now overclocked, so my numbers were fine.

Anyway, small things aside, the soundness of the argument stands. When bitcoins price remains relatively stable, difficulty does not increase that much, and mining is superior to buying coins. When bitcoins price explodes obviously having bought a coin the day before is better. But no one can know the future.

As Litt says, he wishes he had bought a buncha coins instead of buying mining equipment. Why didn't he? Because he couldn't predict the future. No one knew that bitcoin was going to explode from $1 / BTC to $11 / BTC in a matter of 2 months. It might have stagnated forever. And it might continue to rise or it might plummet like a rock, in which case buying coins is far far far stupider than buying mining equipment.


My only argument from the start has been that you don't know the future, so speculating that one approach or the other is better is irrelevant. You can guess as to what you think might happen, and plan accordingly, and that's the best you can do. But don't tell people that it will definitely be this way or that.

If this is your first time then I can guarantee you that you can cut the cost of that $2600 Mining Rig in half and still get the same hashrate.

I guess some people are missing the point that some people may just REALLY enjoy the mining experience. That variable cannot be measured. I always tell people I can't get into golf because I would spend too much time and money on something I don't fine value in. Some guys look at me crosseyed because they enjoy it so much. Same with skiing, water or snow or any other damn sport for that matter.

Personally I view people trying to convince others to just by bitcoins similarly to people who try to get others to believe their religion. They feel better if they can win converts. They're burned out on mining....maybe weren't that good with hardware or would just really rather be doing something else. Which is perfectly fine. But on and on with it these people!!

The way I do math, you can't lose mining. And I will not debate it with anyone.

Cheers

I dont have enough cash to warrant investing in something that might possibly be worth zilch tomorrow.

I know that my computer hardware will still be worth something tomorrow.

I would say that the speculating "economists" are the rubes.