How long will it take to mine 1,000 Bitcoins at 3Ghps?

[Cablesaurus]

So I'm just doing some figuring in my head, some folks are pretty good around here at projecting possible difficulty increases.

Who can come up with the best estimate as to how long it will take to mine, say, 1,000 Bitcoins @ 3GHash, factoring in realistic assumptions about difficulty?

[1714063470]

1714063470

[bulanula]

What rig do you plan on using for this goal ?

[MrGaSp]

About 80 days with my quick estimating...are you actually going to do this?

I figured difficulty at 60% increase next, then 80% and then 90%..because at that point, people may be getting worrisome..

I guess you could figure when AMD will be releasing their new series of boards...and assume a huge jump that difficulty span.. =P

80 days, if not more.

[proudhon]

You actually probably won't ever reach 1000BTCs at 3Ghash/s at the rate the difficulty is increasing.  If you started yesterday, and we assume a 40% difficulty increase at each difficulty, and about 10 days per difficulty period, then by the end of the year you'd have ~670BTCs and you wouldn't ever reach 671 because you'd be bringing in like millionths of a bitcoin each period.  In order for you to make it to 1000BTCs with 3Ghash/s the difficulty would have to increase at an average of some value less than ~25%.

[ribuck]

Suppose difficulty increases by 40% every 10 days. Then in each successive 10-day period your rig will mine this many coins:

190 + 137 + 98 + 70 + 50 + 36 + 25 + 18 + 13 + 9 + 7 + 5 + 3 + 2 + ...

After 140 days you will only have mined 663 coins, and they'll be coming in mighty slowly after that and you would never reach 1000.

Having said that, I did a similar calculation when I bought my 5870 in November last year, and I thought that it would never earn back its cost. Luckily, the difficulty increase turned out to be much lower than I had anticipated in the weeks after I bought the GPU card, and it turned out to be a good purchase.

But recently the difficulty seems to be increasing faster again.

[Raulo]

Assuming 40% increase in this round and another, then 30% then 25% and then 15% afterwards per each round (which is realistically optimistic I would say), you will mine 1000 BTC after 233 days). At day 233, difficulty of 4.07994e+06 results in  0.739591 BTC per day.

If the BTC exchange rate keeps rising, the difficulties will increase so rapidly that probably you will never mine 1000 BTC.

[serial]

Basically, your 3 Ghash/s rig will wear out before it ever mines 1,000 BTC.

The most it will ever do will be more like 427 BTC if exponential price increase continues, up to 494 BTC if prices stabilize. Neither of those things will happen, but that gives you a range you can reasonably expect.

This sounds rather apocalyptic, if a 3Gh/s rig cant make a single bitcoin in a day by by January why would anyone do it? The electricity cost is higher than that.

[rezin777]

This sounds rather apocalyptic, if a 3Gh/s rig cant make a single bitcoin in a day by by January why would anyone do it? The electricity cost is higher than that.

Bitcoin exchange rate increases along with a desire to secure the network from attack?

[ribuck]

if a 3Gh/s rig cant make a single bitcoin in a day by by January why would anyone do it? The electricity cost is higher than that.

There's a possibility that the Bitcoin value will be much higher by January.

When I bought my 5870 GPU card six months ago it could get a block every couple of days. But the difficulty rose very quickly, and now it only mines one block every 25 days on average. But Bitcoins were worth $0.25 when I bought the card and now they're worth 20 times as much. Provided the Bitcoin price holds, my GPU card is actually more profitable now than it was six months ago.

[serial]

There's a possibility that the Bitcoin value will be much higher by January.

When I bought my 5870 GPU card six months ago it could get a block every couple of days. But the difficulty rose very quickly, and now it only mines one block every 25 days on average. But Bitcoins were worth $0.25 when I bought the card and now they're worth 20 times as much. Provided the Bitcoin price holds, my GPU card is actually more profitable now than it was six months ago.

Ok but if it takes 25 days per block now, and that difficulty is going to increase 60-80-90-90-90% then if I'm reading that right its going to take 136 days per block in just a single month's time. Then 259 days a block at 40 days, and 493 days at 50 days to 938 days in just 2 months time, 6424 days in 3 months, 44143 days in 4 months, 302781 in 5, 2076778 in 6, and 14244621 by January or 39,026 years per block .

I guess what I'm asking, is if bitcoin needs new blood like myself in order to increase in acceptance and whatnot, but if there's no conceivable way I can build a machine right now that can pay itself off and cover the electric bill what incentive do I have to even get started in this venture.

The way I have it worked out right now I've been running numbers day and night for the last 8 days and the most efficient machine I can build economically is going to be 1500 dollars at 1Gh/s, that's around 8 days per block or ~31 days per gpu times 4 at the current difficulty.

The only way I see to even contribute is to join a mining pool, but at 5 btc/day and falling the prospect of it paying itself off at the 60 day mark as I had originally calculated it, seems bleak.

Bitcoin exchange rate increases along with a desire to secure the network from attack?

The value of the bitcoins would have to follow suit to the difficulty increases wouldn't they? So say today's market value of 5.60 on the same 60-80-90-90-90-> Would have bitcoins
5.60.. 9.. 16.. 30.. 58.. 110.. 210.. 399.. 758.. 1441.. 2739.. 5204.. 9888.. 18787.. 35696.. 67823.. 128863.. 244841.. 465198.. 883876.. 1679365.. 3190795.. 6062510 = January

So I guess what I'm saying is if the bitcoins rise exponentially in difficulty over the next few months, that means either the value of the bitcoins themselves have to rise at a similar exponential rate since by January using that formula its going to take 39 thousand years to do what takes 25 days now.

If there's flaw in my numbers please let me know I'm just seeing this as a bad omen that the difficulty and btc values are on such an exponential rise, it can only mean the value of everyone who has been hoarding coins is going up at an undeserving rate and will collapse the market when they decide to sell.

[eleuthria]

The problem is people are making projections based on current exponential growth.  That WILL stop, but we can't be sure when.  For a couple more difficulty increases, it will probably remain growing at the recent rate.  If the price of BTC continues to grow roughly in line with the difficulty, the growth will likely continue.

It will NOT continue for a year, unless in the next year something revolutionary happens with processing power.  There simply aren't enough chips around to sustain an exponential growth that is compounded every 2 weeks.

However, if mining becomes complete unprofitable, people will quit, the difficulty will go down.  There will be an equilibrium where it is profitable for some, but others can't compete with either the power efficiency of some miners, or the electricity rates.

[serial]

There are still lots of reasons to operate mining rigs. For one, it still may be one of the best ways of converting local currency into BTC while remaining anonymous. There are probably a few out there who still place a premium on such privacy and will gladly pay the overhead compared to buying BTC on the exchanges or exposing themselves to counter-parties in other ways. For others, the hardware is already just sitting around and it can be used to support the network during what would otherwise be idle time. If it is just electricity that is being paid for such a rig could still pay for its own operation at least until mid 2012 if it is a reasonably efficient rig at average rates, and the amount it can make in the mean time could offset its electric cost of operation much farther into the future.

So, does that mean people won't want to mine? I think there will still be plenty of miners and the network will continue to grow despite all this. Although, it will likely mean that people will want to find more efficient ways to build and operate mining nodes. By my reckoning the Bitcoin network has just left Folding@Home in the dust as the world's largest distributed super-computer and by the end of the year I don't even think a concerted bot-net attack could touch it.

Essentially then what you're saying is the only people who will be able to afford to mine are those who were going to purchase the hardware anyway and only need to make a few cents over the devaluation of the hardware plus electric cost.

That puts me out of the project, the hardware has to earn enough to pay itself off entirely within a few months or it would be a terrible business decision on my part and would make more sense to just purchase the bitcoins and wait for the exponential growth to make them worth selling.

I really wanted to contribute over time and earn residual income but it looks like the smart money would be to just buy and sit on bitcoins like a day trader.

[rezin777]

This sounds rather apocalyptic, if a 3Gh/s rig cant make a single bitcoin in a day by by January why would anyone do it? The electricity cost is higher than that.

Bitcoin exchange rate increases along with a desire to secure the network from attack?

It is the other way around. A high exchange rate drives network expansion.

Of course, that's what I meant.

I'll rephrase. People will mine in January if the exchange rate increases, and also if they desire to secure the network from attack (to secure coins they already own).

[SgtSpike]

You guys should check out my spreadsheet (the "mine or invest" link in sig).  It'll answer all your questions.  And unless you buy a stupidly-expensive computer vs how much MH/s it puts out, you can easily make it pay for itself at current projected difficulty and projected bitcoin valuation changes.

[SgtSpike]

You guys should check out my spreadsheet (the "mine or invest" link in sig).  It'll answer all your questions.  And unless you buy a stupidly-expensive computer vs how much MH/s it puts out, you can easily make it pay for itself at current projected difficulty and projected bitcoin valuation changes.

Yes, a mining rig could still easily pay for itself at this point, but it is unlikely that it will pay you more BTC than you could have bought with the same amount of cash.

Did you check the spreadsheet?  It greatly depends on difficulty and bitcoin value increase, but using difficulty increases of 40%, and BTC value increases of 10% every week, the calculations said you're wrong.  At least, as of a couple of days ago, they did.  You get more money back by investing in mining hardware than directly in coin.

[SgtSpike]

Lol.  I've found my fair share of mistakes in my own formulas as well, so I can't fault anyone else for having errors in theirs.  Glad it's a better outlook and not a worse one though. 

[ribuck]

The viability of mining basically boils down to this:

Price increases faster than difficulty => profit!
Difficulty increases faster than price => ruin.

[Raulo]

Price increases faster than difficulty => profit!
Difficulty increases faster than price => ruin.

Not really. With constant BTC prices, it is hard to get ruined.

Difficulty increases faster than price =>small profit/no profit
BTC price crash =>moderate loss

It's hardly a ruin because the most you can lose is electricity cost + hardware depreciation.

Buying BTC instead of mining could well bring more profit but is much more risky in case of price crash.

By the way, scenario "price increases faster than difficulty" is impossible long term. Mining is already quite profitable, more price increases would lead only to even more mining activity, hence more difficulty.

[Zibbo]

For some reason people are treating difficulty levels as some sort of force of nature, which will increase exponentially regardless of any other variables. This is crazy. Mining difficulty depends almost entirely on BTC exchange rate, which is really high now compared to the difficulty, making mining insanely profitable (with current difficulty of course). This encourages people to start mining, causing difficulty to rise. Current huge profits obviously will not last, because at some point mining difficulty will reach BTC exchange rate (by difficulty increasing or exchange rate dropping), and a lot of people are forced to quit mining or mine at loss (which I predict not many are willing to do). For example if BTC suddenly dropped to $0.5 tomorrow and stayed there, I guarantee that mining difficulty would drop very quickly.

So if you are planning to mine for an extended period of time, you should concentrate on Mhash/W, because when the time comes when the profit margins are a lot smaller (and that time WILL come, sooner rather than later), you want to be able to ride it out, when people with less cost effective rigs are forced to quit. For example, I get 2.1 Mhash/W, pay 0.12€/kWh, and can use wasted energy for heating 7-8 months of the year (effectively cutting my electric bill in 1/3). When the time comes, which one of us is going to drop out first, making it that much more profitable for the other?

Exponential growth in mining difficulty requires an exponential growth in BTC exchange rate, and if you believe with some certainty that is going to happen, you should forget mining, take all your money, and buy bitcoins as quickly as possible. Investing in bitcoins pays of only if bitcoins increase in value. Investing in a mining rig can pay of with both increasing AND decreasing value, but only if your rig is more cost effective than significant portion of other miners.

[SgtSpike]

According to my spreadsheet... a $3,300 rig @ 2800MH/s would make $1,631 worth of bitcoins in 2 months, even assuming the value of BTC stays at $5.71, and assuming a 28% difficulty rate increase every difficulty bump!  It also accounts for 10% depreciation of the hardware each month, electricity costs, and other items.  As long as you can part it out and get $2,673 for the hardware after two months, you've just pocketed $1600.