I've had my eye on a 7970 for a while and it's about time to upgrade the gfx card in my computer as it's not playing the new games as well as I would like. I know there are a lot of people here trying to sell GPU's since they are not really worth it to mine on anymore.
I'm looking for a card that has a transferable warranty and at least a couple of years left on it so that way in case they have been pushed hard, I'm not out oodles of money.
Can any one recommend any cards and about how much they would run me? Since new ones are about $270 retail, can I expect 30% off? 50% off?
Thanks!
I would think $270 is already about 30% off retail for a 7970. If you can get a new one for that, go for it. |
|
|
|
Hey  I just found a fairly good profitability calculator online -> http://mining.thegenesisblock.com/ <- This one allows you to configure it and it is very configurable. It wants the monthly percentage change in difficulty but the box is automatically populated with 76%... I think it should be somewhere between 100-110 but you can change it. The online calculation is using a 30 day period for calculating reward without a difficulty change midway (or less) so it uses too low a difficulty for 15-20 days per calculation for some periods and it is giving a higher return than reality dictates. I believe it is about 25% on the high side. Also it asks for a predicted rate of rise in Bitcoin value - leave it at 0 and divide your cumulative return by the current Bitcoin value. The value will then be adjusted for power costs and since power rates aren't likely to change considerably over the next few months it will give an accurate depiction of how many coins you can produce over the life of your miner. I think it's great to be able to visualize where the cumulative return begins to decrease  Take care |
|
|
|
It looks like BitFury is delivering Here is my contribution to resolving the bet. Some info and pictures of my units: 400GH unit and 25GH unit. Total hashrate ordered 425GH. They seem to be working better without fans. This is what I got yesterday:  Power consumption is for 16 H-Boards.  I'm still tweaking my units manually. I'd say 380GH from 16 H-boards is possible.   Here is a picture from my pool. Hashrate estimate is based on the previous 45 minutes. Both units are connected to worker "Bebber_1".  The $20k initial price tag on these was way too high as was the $12k for October delivery. Now they're saying $8k for October delivery which is still quite a risky venture IMHO. Take care |
|
|
|
Hi xhomerx10. I'd be interested in the specifics you used in your calculations. A couple of observations. First, difficulty has been going up by more than 2.5%/day and the next increase looks to continue this trend. Secondly, don't forget to factor in the cost of powered USB hubs to operate your satoshi sticks.
Cheers. Hi Creativex, I was calculating a compounded rate on diff and if you take 2.5% compounded daily over 10 days (which is as long as the Bitcoin network took to finish the last 2 sets of 2016 blocks) it works out to just shy of a 30% and I think the last two changes were right around there. I didn't factor in electricity costs - I am merely trying to show the maximum number of coins you can generate. If you can't mine at least as many bitcoins as it would take you to buy it at today's exchange rate, it is simply not worth buying the device. The bitcoins would be a better purchase for your money. Essentially, I calculated the current reward and used a power series to factor in the rate of change of difficulty hence the declining reward. Consider h is hash rate in hashes per second, t is time in seconds, D is difficulty, the probability of generating a block is 1/(2^32*D) and the current block reward is 25 BTC reward = 25*h*t/(2^32*D) now substitute 86400 seconds for the time to represent one day's reward for each subsequent day, increase D by 2.5% (so D*1.025) of the previous day and as you can see we have a convergent power series: 1 + x + x^2 + x^3... whose sum can be calculated using the formula [x^(m-1)-1]/(x-1) for m iterations OR to simplify further, we can use an infinite number of iterations which gives us the sum 1/(1-x) x will have the value 1/1.025 (because D is in the denominator) if we consider the 2.5% per day compound increase so the sum of this power series is 1.025/.025 or 41 Simply multiply the initial reward by 41 to get the total number of coins you can mine in an infinite number of days. So total reward (@ 2.5%/day increase in difficulty) = [h/d] * [(41*25*86400)/2^32] which you can use for any device. If you want to use a different percentage increase, substitute 1/(1-x) for 41 (so for your 3% assumption we would get 34.3) Right now I have 13.6 GH/s in USB miners and my total reward will be 4.26 BTC if the network hash rate continues to increase at this pace. I realize there is actually a step change in difficulty every 2016 blocks but even if I use a 10 day period (which is how long the last 2016 blocks took) and 30% compounding instead, the result is only slightly larger ~ 4.5 BTC. I don't even want to say how much I spent to get them but I can't make an argument without it so... about 22BTC. Now if those 4.5 BTC I can ultimately generate are worth as much as the 22 BTC I spent to get them, then I'm good but imagine what those 22 BTC would have been worth if I had saved them instead ;( Live and learn! Take care PS based on the 2.5% increase, a 2 TH/s miner received 90 days from today would mine 1.65 BTC/day so you're right, at 14k, it's still only worth a wait-and-see! Thanks for showing your work. As we're a BTC based company paying dividends daily in BTC, I think of returns in terms of BTC and leave fiat exchange rates out of the maths whenever possible. Obviously it's not possible to exclude exchange rates entirely from my analysis as many(most) hardware vendors price their products in fiat. In this instance I would use a one time exchange rate conversion which would then become the product's initial cost. A positive ROI would be achieved when that number is exceeded by output from said equipment. Cheers. It is hard to remove exchange rates entirely, especially when trying to factor in electricity costs. Luckily the market is helping us now by making mining hardware obsolete before we can even plug it in! I'm still waiting for my BFL 60Gh/s space heater. I figure it will only be good for keeping my feet warm under the desk this winter or next spring.... Take care |
|
|
|
Hi xhomerx10. I'd be interested in the specifics you used in your calculations. A couple of observations. First, difficulty has been going up by more than 2.5%/day and the next increase looks to continue this trend. Secondly, don't forget to factor in the cost of powered USB hubs to operate your satoshi sticks.
Cheers.
Hi Creativex, I was calculating a compounded rate on diff and if you take 2.5% compounded daily over 10 days (which is as long as the Bitcoin network took to finish the last 2 sets of 2016 blocks) it works out to just shy of a 30% and I think the last two changes were right around there. I didn't factor in electricity costs - I am merely trying to show the maximum number of coins you can generate. If you can't mine at least as many bitcoins as it would take you to buy it at today's exchange rate, it is simply not worth buying the device. The bitcoins would be a better purchase for your money. Essentially, I calculated the current reward and used a power series to factor in the rate of change of difficulty hence the declining reward. Consider h is hash rate in hashes per second, t is time in seconds, D is difficulty, the probability of generating a block is 1/(2^32*D) and the current block reward is 25 BTC reward = 25*h*t/(2^32*D) now substitute 86400 seconds for the time to represent one day's reward for each subsequent day, increase D by 2.5% (so D*1.025) of the previous day and as you can see we have a convergent power series: 1 + x + x^2 + x^3... whose sum can be calculated using the formula [x^(m-1)-1]/(x-1) for m iterations OR to simplify further, we can use an infinite number of iterations which gives us the sum 1/(1-x) x will have the value 1/1.025 (because D is in the denominator) if we consider the 2.5% per day compound increase so the sum of this power series is 1.025/.025 or 41 Simply multiply the initial reward by 41 to get the total number of coins you can mine in an infinite number of days. So total reward (@ 2.5%/day increase in difficulty) = [h/d] * [(41*25*86400)/2^32] which you can use for any device. If you want to use a different percentage increase, substitute 1/(1-x) for 41 (so for your 3% assumption we would get 34.3) Right now I have 13.6 GH/s in USB miners and my total reward will be 4.26 BTC if the network hash rate continues to increase at this pace. I realize there is actually a step change in difficulty every 2016 blocks but even if I use a 10 day period (which is how long the last 2016 blocks took) and 30% compounding instead, the result is only slightly larger ~ 4.5 BTC. I don't even want to say how much I spent to get them but I can't make an argument without it so... about 22BTC. Now if those 4.5 BTC I can ultimately generate are worth as much as the 22 BTC I spent to get them, then I'm good but imagine what those 22 BTC would have been worth if I had saved them instead ;( Live and learn! Take care PS based on the 2.5% increase, a 2 TH/s miner received 90 days from today would mine 1.65 BTC/day so you're right, at 14k, it's still only worth a wait-and-see! |
|
|
|
|
Hey,
I recently used one of the profitability calculators on the internet to see if acquiring some USB erupters was a worthwhile venture. It seemed as though I could theoretically see my money returned. Using that site was lazy I'll admit and it will be my own fault when they are not profitable which will be the case. At any rate, to redeem myself I sat down yesterday and actually figured out conservatively how many coins a given miner will produce in its lifetime in the current climate. The reason I am posting it here is that I think it will be of value to any further acquisitions made by bASIC-MINING. I can't show all the math because I don't know how to display summation notation and powers with this editor.
Assumptions
You will acquire the device the day you do the calculation
Block reward is 25 coins
Rate of network difficulty change 2.5%/day
Calculation
Divide the hashing rate (in hashes per second) of your desired machine by the current difficulty and multiply by 0.0201166 <---- rounded off
Conclusion
A 5 GH/s BFL Jalapeno arriving today will produce a total of 1.6BTC (ever!)
A Bock erupter Blade purchased for 5.79 BTC today will produce a total of 3.977 (ever!) <--- provided you run it over-clocked at 13GH/s
If you don't believe 2.5%/day and would prefer to use 2%/day then you can multiply your result by 1.2
Want to use 1.5%? Then use the factor 1.67
Take care
|
|
|
|
I PM'd him with the Alberta one but I just realized now reading your post that it was a 30G. I thought it was a 50! Seems kinds pricey now. Take care |
|
|
|
Live: http://www.basic-mining.com--------------------------------------------------------------------------------------- Drop comments in mailbox: demzie@basic-mining.comHave any FAQ idea's? Drop FAQ entries in mailbox: faq@basic-mining.com--------------------------------------------------------------------------------------- There are still a few pages i want todo but those need input from the management - About bASIC-Mining (Written as a story) - Our Goals and Future - Portfolio page (Just great pictures of bASIC-Mining Hardware collection) - Maybe more pages based on input from you or the management ofcourse The todo pages will come with time and planned urgency as management will decide --------------------------------------------------------------------------------------- Very nice! |
|
|
|
Thank you. It's going to be hot this week(for MI) and I don't think I can/should put off the purchase of cooling gear any longer so I'm going to look into that this evening. I believe we could see some benefit from deploying some strategically placed high velocity industrial 120v fans so I'm going to look into pricing several 14" and 20" units. Cheap retail box fans are next to useless as they don't move enough air. Cheers. Hey Creative, You're doing a great job steering your company in the right direction so I almost hate to bring this up but I think it might be prudent to shut down the GPU miners if heat is an issue. It's probably adding 400+ watts of heat (which is the equivalent of a BFL single I think) and generating relatively little if any profit. If the heat is going to jeopardize the operation of your next-gen miners, taking 400 watts out of the equation might be as good an option as adding a fan. I've been experimenting with middlecoin and my 1100 kH/s rig and I would be hard-pressed to say it is profitable. My average electricity cost is $0.18 per kWh though - we are on a time of use system so it is possible to profit if I only run it on weekends and at night but that would be a pain in the ass for the less-than-50 cents a day profit Take care |
|
|
|
...and now
[2013-08-18 20:52:17] Network diff set to 59.8M
What coin is diff 59.8 million?!? Anybody?
Take care
That one is Litecoin. And the 21.2M one you linked earlier is Novacoin. Ah! I get it now - thanks for all the help! Take care My litecoin client says diff is ~913 and Novacoin ~340 so where does this discrepancy come from? Does h2odysee apply some constant fudge-factor to throw us off the trail of which coin the pool is currently mining? It's how cgminer reports the network diff, middlecoin just follows it. So I am also running a second instance of cgminer for Bitcoin and it reports a network diff of 50.8M - I appreciate your willingness to help but I'm just not understanding. I read somewhere that cgminer reports network diff at 65,000 times the actual diff for other coins. Why? I dono, just seems to be that way. Thanks samfisher. I pointed my miner at a few different scrypt pools and it appears as though the displayed diff in cgminer is difficulty*2^16 though I still can't understand the why. I think I'm going to accept that there is a difference and let it go now At least I can figure out which coin we're mining on middlecoin.com now! Take care Difficulty 1 Litecoin (read: Scrypt) shares are set at 65,535 ("FFFF" or 16^4). Edit: to clear confusion a bit (as to how an even * an event = an odd. It's because technically 16^4 is 65,536 but since the computer reads 0 as the first number we subtract one (65,536-1=65,535). |
|
|
|
...and now
[2013-08-18 20:52:17] Network diff set to 59.8M
What coin is diff 59.8 million?!? Anybody?
Take care
That one is Litecoin. And the 21.2M one you linked earlier is Novacoin. My litecoin client says diff is ~913 and Novacoin ~340 so where does this discrepancy come from? Does h2odysee apply some constant fudge-factor to throw us off the trail of which coin the pool is currently mining? It's how cgminer reports the network diff, middlecoin just follows it. So I am also running a second instance of cgminer for Bitcoin and it reports a network diff of 50.8M - I appreciate your willingness to help but I'm just not understanding. I read somewhere that cgminer reports network diff at 65,000 times the actual diff for other coins. Why? I dono, just seems to be that way. Thanks samfisher. I pointed my miner at a few different scrypt pools and it appears as though the displayed diff in cgminer is difficulty*2^16 though I still can't understand the why. I think I'm going to accept that there is a difference and let it go now At least I can figure out which coin we're mining on middlecoin.com now! Take care |
|
|
|
...and now
[2013-08-18 20:52:17] Network diff set to 59.8M
What coin is diff 59.8 million?!? Anybody?
Take care
That one is Litecoin. And the 21.2M one you linked earlier is Novacoin. My litecoin client says diff is ~913 and Novacoin ~340 so where does this discrepancy come from? Does h2odysee apply some constant fudge-factor to throw us off the trail of which coin the pool is currently mining? It's how cgminer reports the network diff, middlecoin just follows it. So I am also running a second instance of cgminer for Bitcoin and it reports a network diff of 50.8M - I appreciate your willingness to help but I'm just not understanding. |
|
|
|
...and now
[2013-08-18 20:52:17] Network diff set to 59.8M
What coin is diff 59.8 million?!? Anybody?
Take care
That one is Litecoin. And the 21.2M one you linked earlier is Novacoin. My litecoin client says diff is ~913 and Novacoin ~340 so where does this discrepancy come from? Does h2odysee apply some constant fudge-factor to throw us off the trail of which coin the pool is currently mining? |
|
|
|
Something doesn't look right here
2013-08-18 09:29:31] Accepted 70aa8da1 Diff 2.02K/512 GPU 1
[2013-08-18 09:30:07] Accepted edc9a467 Diff 546/512 GPU 0
[2013-08-18 09:30:34] Accepted 9ea2b04a Diff 1.16K/512 GPU 0
[2013-08-18 09:30:44] Stratum connection to pool 0 interrupted
[2013-08-18 09:30:44] Pool 0 stratum share submission failure
[2013-08-18 09:30:49] Network diff set to 42K
[2013-08-18 09:30:49] Pool 0 communication resumed, submitting work
[2013-08-18 09:30:49] Rejected d37718e0 Diff 663/512 GPU 0 (Job '20a9' not fou
d)
[2013-08-18 09:30:49] Stratum from pool 0 requested work restart
[2013-08-18 09:30:51] Stratum from pool 0 detected new block
[2013-08-18 09:31:11] Network diff set to 168K
[2013-08-18 09:31:11] Stratum from pool 0 detected new block
[2013-08-18 09:31:11] Stratum connection to pool 0 interrupted
[2013-08-18 09:31:21] Network diff set to 21.2M
[2013-08-18 09:32:17] Accepted a89dcafb Diff 783/512 GPU 1
[2013-08-18 09:32:58] Accepted b01d5de4 Diff 4.18K/512 GPU 0
Looks like I was mining along with network diff @ 21M (weird number) for a while then we start mining something else for 30 seconds then switch to another coin for 10 seconds and then back to network diff 21M which is a strange value since there is no altcoin at that level other than Namecoin which is not a scrypt-based currency anyway. Thirty seconds? Ten seconds? 21M?? Is this a programming issue or is that a realistic time to mine with the middlecoin pool?
Thanks
Take care
...and now [2013-08-18 20:52:17] Network diff set to 59.8M What coin is diff 59.8 million?!? Anybody? Take care |
|
|
|
This is assuming difficulty continues to double as fast as it does. Well, OK, there's Moore's law and all.
However there are many other likely things that can happen in the future:
* finding places with cheap electricity
* having better energy sources (you could even run a 100% solar power operation, stopping the miners when you don't have your free power if it really is expensive)
* bitcoin price going up
I have investigated the solar power aspect. It is not cheaper than current electricity costs in my area. In fact if you do all of the work of installing the panels yourself, buy them as a skid and don't include any batteries for storage, you're looking at $1.50 per watt (that's watt!) installed. Not if you live near some high tension lines and have a little copper coil setup... well that would be illegal! Take care |
|
|
|
This is assuming difficulty continues to double as fast as it does. Well, OK, there's Moore's law and all. This isn't related to Moore's law in any way. We simply know that a lot of ASIC mining devices are being manufactured, and many of them will be more efficient than the current generation. Extrapolation simply gives you an idea how fast this process is. IMHO, It is directly related to both Moore's first and second law. Inititally, there was more demand for FPGAs/ASICs and now there is more demand for efficiency and speed in these. Speeds have and will double every 18 months to 2 years and price will reflect that. Perhaps the network will even reach an equilibrium at some point - provided the concept of a digital currency can adapt. Take care |
|
|
|
|
Something doesn't look right here
2013-08-18 09:29:31] Accepted 70aa8da1 Diff 2.02K/512 GPU 1
[2013-08-18 09:30:07] Accepted edc9a467 Diff 546/512 GPU 0
[2013-08-18 09:30:34] Accepted 9ea2b04a Diff 1.16K/512 GPU 0
[2013-08-18 09:30:44] Stratum connection to pool 0 interrupted
[2013-08-18 09:30:44] Pool 0 stratum share submission failure
[2013-08-18 09:30:49] Network diff set to 42K
[2013-08-18 09:30:49] Pool 0 communication resumed, submitting work
[2013-08-18 09:30:49] Rejected d37718e0 Diff 663/512 GPU 0 (Job '20a9' not fou
d)
[2013-08-18 09:30:49] Stratum from pool 0 requested work restart
[2013-08-18 09:30:51] Stratum from pool 0 detected new block
[2013-08-18 09:31:11] Network diff set to 168K
[2013-08-18 09:31:11] Stratum from pool 0 detected new block
[2013-08-18 09:31:11] Stratum connection to pool 0 interrupted
[2013-08-18 09:31:21] Network diff set to 21.2M
[2013-08-18 09:32:17] Accepted a89dcafb Diff 783/512 GPU 1
[2013-08-18 09:32:58] Accepted b01d5de4 Diff 4.18K/512 GPU 0
Looks like I was mining along with network diff @ 21M (weird number) for a while then we start mining something else for 30 seconds then switch to another coin for 10 seconds and then back to network diff 21M which is a strange value since there is no altcoin at that level other than Namecoin which is not a scrypt-based currency anyway. Thirty seconds? Ten seconds? 21M?? Is this a programming issue or is that a realistic time to mine with the middlecoin pool?
Thanks
Take care
|
|
|
|
hi! is this a good pool? how long after submitting shares does my address show up on the front stat page? can i use one single btc address with all of my miners? thanks for the info It's a great pool and a wonderful piece of work. After a couple of minutes your addy should appear (share diff is 512) Yes, you can! Take care |
|
|
|
The difficulty is killing everyone, including stocks such as bASIC. I read earlier today that any BFL device that arrives later than September 1st will be unprofitable with the current difficulty rising trend.
Unfortunately that isn't just affecting just BFL customers, it affects bASIC and all other securities or similar. I think the way they should be focusing on now, is using sites like Middlecoin, or multipool (when they implement their auto-exchange feature) and using GPU's for mining. I know they're highly in-efficient, but they can generate more profits mining than ASICs can, I never thought I'd say that, but I guess things change, aka the difficulty of bitcoin mining. Hi guitarplink. I find this line of thinking questionable. Difficulty will go up until mining is no longer profitable for those with inefficient gear and/or high overhead. It cannot go up far beyond this point because many(most) will stop mining which will cause difficulty to fall. While the pre-order madness has created a market distorting force, it will not last forever. If you skim the forums here you will note a growing resistance to the risky pre-order model. bASIC-Mining is well capitalized and IMO well positioned to weather this storm. We're mining with GPUs now on Middlecoin...they earn next to nothing. They're nowhere near as profitable as ASICs. This will not change in the near future. Cheers. Before I continue, I'm not trying to troll or anything, I'm just trying to understand. If the high difficulty puts other ASIC miners out of the game, what is stopping bASIC mining from losing money by mining as well? Do they operate on very low power costs, or keep mining even when they lose money, or whats up? And when the difficulty becomes too high for ASICs to be profitable, yes people will take them offline. However at the next difficulty change there is a good possibility that the difficulty will make it profitable again, until the next difficulty change, and it will continue like that. Like, does bASIC have a plan to combat this all? Hey Guitarplinker, Creativex has more important things to do than allay your fears. He's not a psychologist, he's an entrepreneur with a plan! Clearly you have not done any research before asking your questions. Read something first so you can ask more informed questions. Take care I'd love to do more research, but I cannot find any more information on bASIC mining, do you know of a place I can go to read all about it, it's goals, and how it works? No you wouldn't. You're line of questioning is based on your misunderstanding of the technology. bASIC-mining's goal is to produce bitcoins and increase it's production of bitcoins. It works by investing in the latest technology available to mine bitcoins. It also works by having a dedicated, resourceful and genius founder. Please stop wasting his time. Take care |
|
|
|
The difficulty is killing everyone, including stocks such as bASIC. I read earlier today that any BFL device that arrives later than September 1st will be unprofitable with the current difficulty rising trend.
Unfortunately that isn't just affecting just BFL customers, it affects bASIC and all other securities or similar. I think the way they should be focusing on now, is using sites like Middlecoin, or multipool (when they implement their auto-exchange feature) and using GPU's for mining. I know they're highly in-efficient, but they can generate more profits mining than ASICs can, I never thought I'd say that, but I guess things change, aka the difficulty of bitcoin mining. Hi guitarplink. I find this line of thinking questionable. Difficulty will go up until mining is no longer profitable for those with inefficient gear and/or high overhead. It cannot go up far beyond this point because many(most) will stop mining which will cause difficulty to fall. While the pre-order madness has created a market distorting force, it will not last forever. If you skim the forums here you will note a growing resistance to the risky pre-order model. bASIC-Mining is well capitalized and IMO well positioned to weather this storm. We're mining with GPUs now on Middlecoin...they earn next to nothing. They're nowhere near as profitable as ASICs. This will not change in the near future. Cheers. Before I continue, I'm not trying to troll or anything, I'm just trying to understand. If the high difficulty puts other ASIC miners out of the game, what is stopping bASIC mining from losing money by mining as well? Do they operate on very low power costs, or keep mining even when they lose money, or whats up? And when the difficulty becomes too high for ASICs to be profitable, yes people will take them offline. However at the next difficulty change there is a good possibility that the difficulty will make it profitable again, until the next difficulty change, and it will continue like that. Like, does bASIC have a plan to combat this all? Hey Guitarplinker, Creativex has more important things to do than allay your fears. He's not a psychologist, he's an entrepreneur with a plan! Clearly you have not done any research before asking your questions. Read something first so you can ask more informed questions. Take care |
|
|
|
|
Show Posts
