• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Bitcoin Discussion Thread Part 2

Got my pair of 7970s. Put em to mine using diablominer @1Ghz clock. Results are impressive:

7970c.jpg
 
^
Cool, that's what one of my triple 5870 rigs do. I wonder how much $ you would save on power consumption?

For the price of two 7970s though, you could get 6-7 5870s!
 
bitcoins got a nice bump from "the good wife" yesterday. The entire plot of the show revolved around bitcoins as an illegal electronic currency and the attempt to find the founder so they could prosecute him....
 
bitcoins got a nice bump from "the good wife" yesterday. The entire plot of the show revolved around bitcoins as an illegal electronic currency and the attempt to find the founder so they could prosecute him....

No shit? Did they actually call them bitcoins on the show?

Wow, just did some Googling and read a few articles. The name of the episode was actually "Bitcoin for Dummies". Ha!
 
Last edited:
Huh, so you managed to get the memory clocks low. I wasn't having any success at all with that. I'll have to make sure my Afterburner is set with overclocking mode 2. Sounds like that might be the trick.

Hmm, I can't get mine to go below 685 on the memory. I keep killing Afterburner and reopening it and it never opens up below 685.
 
Hmm, I can't get mine to go below 685 on the memory. I keep killing Afterburner and reopening it and it never opens up below 685.

That's an Afterburner/Videocard bug. You can manually set lower it by dragging the slider down, hitting apply, and switching to another GPU but it will never open up at a value below the stock threshold unless you flash the bios.
 
Just started mining the last few days and I have lots of questions, trying to find answers on my own but things are confusing at times.

From what I'm gathering here I'm better off running DiabloMiner (I have two unlocked 6950s at 900+Mhz core) than the default GUIMiner core, and I can greatly reduce power consumption (thereby increasing profitability) by dropping memory speed and possibly core voltage on my cards? Is this process detailed anywhere? I've never tried to undervolt a GPU before, and I imagine it's a pretty delicate balancing act, especially while they're OCed.

I've already got the latest Afterburner beta installed and I'm currently hitting around 1000Mhash/sec from the pair... though I'm not 100% on that number since it varies considerably depending where I'm finding it (GUIMiner, bitclockers.com data, etc...)

Electricity's 12c/KWh here, so I'm not sure if this is really worth it in the long term, but it's an interesting project. Seeing about 450w draw from my UPS without the monitor on, both cards mining.

Assuming I can get around 1200Mhash/sec am I better off solo mining and getting all the profits? From what I'm reading solving a block The machine can easily run 24 hours. Assuming a pool with no fees and minimal pool hopping, is there any real disadvantage to pool based mining vs. solo?
 
Last edited:
at 1200mhash you definitely need to be in a pool. Soloing's not really feasible unless you're pushing 20+ghash.
 
From what I'm gathering here I'm better off running DiabloMiner (I have two unlocked 6950s at 900+Mhz core) than the default GUIMiner core, and I can greatly reduce power consumption (thereby increasing profitability) by dropping memory speed and possibly core voltage on my cards? Is this process detailed anywhere? I've never tried to undervolt a GPU before, and I imagine it's a pretty delicate balancing act, especially while they're OCed.

For a 69XX, I'd go with cgminer. It's got a crapton of useful features. All in the miner client, you can control your cards' core/memory speeds, fan speeds (with automatic fan throttling based on a target core temperature), various multipool options (failover, round robin, load balance, etc.), and a ton of other stuff.

The only reason I use DiabloMiner on my 7970 is that it's currently the only client that supports it.

Assuming I can get around 1200Mhash/sec am I better off solo mining and getting all the profits? From what I'm reading solving a block The machine can easily run 24 hours. Assuming a pool with no fees and minimal pool hopping, is there any real disadvantage to pool based mining vs. solo?

Well, with two unlocked 6950s with a decent overclock, you'll be pulling low-mid 800s Mhash/sec.

But no, even at much more hashing power than that, you'll still want to stick with pool mining. At 850 MH/s, you would average finding a block roughly every 2.5 months. I don't think I would even consider solo unless I was over 5 GH/s.
 
Just started mining the last few days and I have lots of questions, trying to find answers on my own but things are confusing at times.

From what I'm gathering here I'm better off running DiabloMiner (I have two unlocked 6950s at 900+Mhz core) than the default GUIMiner core, and I can greatly reduce power consumption (thereby increasing profitability) by dropping memory speed and possibly core voltage on my cards? Is this process detailed anywhere? I've never tried to undervolt a GPU before, and I imagine it's a pretty delicate balancing act, especially while they're OCed.

I've already got the latest Afterburner beta installed and I'm currently hitting around 1000Mhash/sec from the pair... though I'm not 100% on that number since it varies considerably depending where I'm finding it (GUIMiner, bitclockers.com data, etc...)

Electricity's 12c/KWh here, so I'm not sure if this is really worth it in the long term, but it's an interesting project. Seeing about 450w draw from my UPS without the monitor on, both cards mining.

Assuming I can get around 1200Mhash/sec am I better off solo mining and getting all the profits? From what I'm reading solving a block The machine can easily run 24 hours. Assuming a pool with no fees and minimal pool hopping, is there any real disadvantage to pool based mining vs. solo?

I run cgminer for my 6950s but Diablominer is the newest one and I'm not sure if it's that much better. I am too lazy to rewrite all my scripts and configuration files and scheduled tasks and failover solutions for 12 cards on multiple machines that have to be controlled remotely to switch over to Diablominer though. For anyone who has tried both, have you gotten better megahashes with Diablominer vs cgminer? I know that cgminer was a huge leap over Phoenix. Or is it just the kernel? cgminer is pretty amazing with all the command line features and options and real-time control over your individual cards and workers and also failover pools, etc for if your pool dies it will automatically switch to another. Like Parja, I would recommend cgminer.

Dropping memory speed supposedly reduces your power draw but again, that's one of those things that I've never seen really proven. I have a kill-a-watt on one my 1300MH/s machines and memory clocks @ 1100Mhz vs 400MHz seems to have minimal to no effect on the draw at the wall.

For the cores, it used to be that people wanted to overclock them as fast as possible to get the most hashes out of them and that required bumping up the core voltage. There is, just like in normal overclocking of CPUs, a sweet spot where the power draw to clock speed curve hits the optimal point before returns on geometric increases in power draw and heat are not worth it but this requires a ton of tinkering. It used to be very important to get the most speed of your GPUs because you were fighting the rising difficulty so you wanted to mine as much as possible as fast as possible before difficulty went up. Difficulty is going back up due to prices going back up and people joining in again so I would recommend you not downclock or downvolt but find a good sweet spot for stability, heat, and performance.

Two unlocked 6950s @ 900MHz should not pull 1000MH/s. It should be closer to 850MH/s. Locked ones will do about 700MH/s. I have tried almost all the cards out there in various configurations. Right now I settled on ten 5870s and two 6950s. Used to have 15 but sold 3 when the crash happened.

To get 1200MH/s from unlocked 6950s, you would need three of them. I don't know where you get the impression that you would solve a block in 24 hours. Solo mining at this point with this difficulty means that with average luck, you could go MONTHS without solving a single block. Don't bother solo mining at anything less than 10,000 MH/s. Simplecoin.us is run by a [H]ardforums member and that pool was regularily around 20,000 MH/s and it could go 2 weeks without solving a block at some time.

The disadvantage of running on a pool depends on the type of pool. Some no fee pools rely purely on luck and if the pool is stuck on a block from hell, you could go weeks without getting any dividends...or your machine could go down or be on downtime during a crucial mining period where most shares that count occur before a block is solved and you will get less coins when they are distributed. Some other pools pay per share so it doesn't matter when you mine or not and it doesn't matter what the luck is. You get paid per share of work that you do. These pools will charge small fees (like BTCGuild.com switched to). I prefer this model personally. It was too unpredictable and up/down when relying purely on luck and timing.

At this point in time, I get about 3 BTC/day with about 4000 MH/s for all my machines on a PPS pool (BTC Guild). I have been mining on various pools since June and in all that time, my workers have never been the one to find a block. To put that in perspective, if I had been solo mining, my luck would have been equivalent to mining for over a half a year and gaining a big fat zero BTC.

Incidentally, my trusty A05N became a stripped mining machine in the basement. A little sad, but it is what it is. The layout wasn't conducive to my current system. Silent-circuit are you mining with your A05N? It gets way too hot in there unless the door is off, even with the top hole cover.
 
Last edited:
Somehow missed the end of a sentence -- I never thought I could solve a block in 24 hours, just was saying I could run the machine 24/7. At that rate I think I've seen estimates it'd take about 2 months to solve a block, maybe more maybe less depending on luck, at current difficulty?

Currently in a A05B (not even an NB), GPUs sit around 80C loaded. 55% fan speed. I'll be switching to a BitFenix Survivor in a few days, however, which should be better for cooling.
 
Somehow missed the end of a sentence -- I never thought I could solve a block in 24 hours, just was saying I could run the machine 24/7. At that rate I think I've seen estimates it'd take about 2 months to solve a block, maybe more maybe less depending on luck, at current difficulty?

Currently in a A05B (not even an NB), GPUs sit around 80C loaded. 55% fan speed. I'll be switching to a BitFenix Survivor in a few days, however, which should be better for cooling.

Solo mining is too much of a risk. The best likely scenario would be finding a block once every two months meaning you get 50 BTC plus a tiny, tiny residual from future transaction fees done on those coins. That works out to be 0.64 BTC per day which is comparable to pool mining.

The more likely scenario is that you go 3-4 months or longer without solving a block. As it's dependent on luck, you may not ever get anything at all. With pool mining in a large pool (one over 1000GH/s), you would get paid every day regardless.
 
Nice primer Spare-Flair

Hey guys, I want to shamelessly shill for the simplecoin.us pool.

I was a big BTCGuild partisan until one day they converted to a 5%fee pool without so much as a "by your leave". No communication, no warning, no nothing. It was right around the crash, I get it that they have to pay the bills, but 5% is a bit steep and they could have said something.

Anyway, simplecoin.us , 1% fee to keep the servers running, operator is a tenured [H] member and is responsive on the bitcointalk forums.

We are a small pool, so variance is an issue, but you do not want to be in a big pool if you believe in bitcoin. Centralization is BAD. Deepbit is almost 50% of the hash and this is inherently dangerous to the project.

Anyway, there you go, just a quick plug for my pool...thanks for listening.
 
Nice primer Spare-Flair

Hey guys, I want to shamelessly shill for the simplecoin.us pool.

I was a big BTCGuild partisan until one day they converted to a 5%fee pool without so much as a "by your leave". No communication, no warning, no nothing. It was right around the crash, I get it that they have to pay the bills, but 5% is a bit steep and they could have said something.

Anyway, simplecoin.us , 1% fee to keep the servers running, operator is a tenured [H] member and is responsive on the bitcointalk forums.

We are a small pool, so variance is an issue, but you do not want to be in a big pool if you believe in bitcoin. Centralization is BAD. Deepbit is almost 50% of the hash and this is inherently dangerous to the project.

Anyway, there you go, just a quick plug for my pool...thanks for listening.

I really loved Simplecoin and mined dutifully there for half a year but the variance was too great for me. I'm still #2 in lifetime shares according to the pool but it looks like that stat is not updating.

One too many blocks from hell scared me away unfortunately because of the low hashing power. I was making way less BTC than I could have been on a steady pool. If the pool was bigger, I would be back in no time. I tried SMPPS but then I ended up getting paid once every 2 weeks there. With BTCGuild I get paid every day.

I notice it's almost up to 80MH/s today. Is it steady? That scoinuser guy fluctuates a lot. If he was mining to his max speed most of the time, I would be back on Simplecoin but when it dips to the 20MH/s range, the varience is too unpredictable based on luck.
 
Last edited:
Variance is an issue, that is why I am trying to drum up more hash.

The operator is looking into integrating as a node in p2pool, if it works out that would help also.

It is sort of a catch 22; small pool-->high variance-->scares miners away-->small pool. If we could get some more hash going on we could reduce variance.

I have been happy with the PPLNS system there, but I do not need paid on a schedule.
 
I may try other pools (I've seen some people talking about Eligius?) but for the moment I guess bitclockers is doing well enough for me.
 
I tried mining on Eligius, but in the last couple of days, my miners have sporatically been unresponsive for an hour or two at a time. I switched back to ARSBitcoin, IMO, this is the best pool out there. It isn't too small, or too big and has always been reliable for me.

I used to mine on Simplecoin.us, but with the variance and the couple of long long blocks, I swayed back to ARSBitcoin.
 
Just a note, of course folks are free to do whatever they wish, but I urge anyone tempted to hash for the Eligius pool to look into it carefully. There has been a lot of chatter on the forums accusing the operator, one "Luke Jr." of using his pool's hash to attack other cryptocurrencies.

Sethmo, yeah, I'm down with ARSB, I'm glad folks realize that too big is a problem.
 
nah you should come back and gamble with us for a bit :D

scoinuser and slinky 234 online at the moment, 90 GH in the PPLNS pool alone
 
Last edited:
I dont see much mention of the 6870, is this a viable card for me to try out bitcoining? Reading through a good portion of this thread (and having read about it throughout the past couple years) has piqued my interest.
 
I dont see much mention of the 6870, is this a viable card for me to try out bitcoining? Reading through a good portion of this thread (and having read about it throughout the past couple years) has piqued my interest.

viable? it will hash, it all comes down to hashrate, power consumption and your cost for electricity, there are card hashrate charts out there out there...1.4GH/s is roughly 1 BTC/day at present, get out your power bill and do the math
 
viable? it will hash, it all comes down to hashrate, power consumption and your cost for electricity, there are card hashrate charts out there out there...1.4GH/s is roughly 1 BTC/day at present, get out your power bill and do the math

hmmm thanks for pointing me in the right direction, tried it out:

250watts at 17 cts per kwh is just about $1 a day in electricity cost. At a Rate of 270 Mhash it would take me 5.2 days to get one BTC.
 
pretty marginal operation for you, that is fairly expensive electricity
 
hmmm thanks for pointing me in the right direction, tried it out:

250watts at 17 cts per kwh is just about $1 a day in electricity cost. At a Rate of 270 Mhash it would take me 5.2 days to get one BTC.

That is extremely expensive electricity. For you, it works out to $5.2 in Electricity for 1 BTC ($6-$7 right now). With luck you will make $1. With bad luck, you will lose money.

I pay $0.07/ KWh for electricity. I have approximately 4200MH/s across all miners using 2500W = <$5 in electricity per day. Selling @ $7 means $21/day = ~$16/day profit
 
Last edited:
pretty marginal operation for you, that is fairly expensive electricity

Whoops made a calculation error, I actually pay 13 cts per kwh. (how does that compare?)

So that makes it $4 for 5.2 days of running and 78 cts per day.
 
That makes it sound a little better. lol, the last place I lived had electricity included in my rent. Should've taken advantage, haha.
 
You are making about 48 cents a day in profit. :cool:

HAHA
That sounds like the big bucks to me! In fact, i probably made that much typing this message while at work... (not quite, but you get the idea).

Possibly adding a 2nd card would increase the margins as the system power level would stay equal...
 
doesn't change it much, a lot of us are paying over an order of magnitude less for delivered electrons

my cost/KWh is similar to flair's $0.0664
 
Hmm so:

410watts at 13 cts per kwh is just about $1.28 a day in electricity cost. At a Rate of 540 Mhash (for the extra card) it would take me 2.6 days to get one BTC.

At $6.5 per BTC that is $2.50 a day.

Margins went from .78/1.25=.624
~37.6% profit margin (1x 6870)

To 1.28/2.5=.512
~48.8% profit margin (2x 6870)

Lol, so it matters significantly percentage wise, but dollar for dollar its turning a little into not much more.

Lol it'd take me 117 days of 24hr running to pay off a 2nd 6870...
 
none of my comments are to discourage you...I believe in bitcoin and think they will be worth far more in the future

also, what is the temperature outside where you are? a free 500W heater is nothing to sniff at
 
HAHA
That sounds like the big bucks to me! In fact, i probably made that much typing this message while at work... (not quite, but you get the idea).

Possibly adding a 2nd card would increase the margins as the system power level would stay equal...

Measured power consumption one 6870 card
System in IDLE = 173W
System Wattage with GPU in FULL Stress = 295W
Difference (GPU load) = 122 W
Add average IDLE wattage ~ 19W (ATI specified)
Subjective obtained GPU power consumption = ~ 141 Watts

Measured power consumption two 6870 cards
System in IDLE = 184W
System Wattage with GPU in FULL Stress = 421W
Difference (GPU load) = 237W
Add average IDLE wattage x2 ~ 19W (ATI specified)
Subjective obtained GPU power consumption = ~ 275 Watts

Adding another card is 30% increase in power consumption. If you really are @ 250W, then it'd be around 325W for your system = $1.01/day @ $0.13/KWh but you would make 1 BTC ($6.50) in 2.6 days or $1.50/day in profit.
 
Last edited:
Measured power consumption one 6870 card
System in IDLE = 173W
System Wattage with GPU in FULL Stress = 295W
Difference (GPU load) = 122 W
Add average IDLE wattage ~ 19W (ATI specified)
Subjective obtained GPU power consumption = ~ 141 Watts

Measured power consumption two 6870 cards
System in IDLE = 184W
System Wattage with GPU in FULL Stress = 421W
Difference (GPU load) = 237W
Add average IDLE wattage x2 ~ 19W (ATI specified)
Subjective obtained GPU power consumption = ~ 275 Watts

Adding another card is 30% increase in power consumption. If you really are @ 250W, then it'd be around 325W for your system = $1.01/day @ $0.13/KWh

Ah those numbers look more accurate than the ones I found. But mine are decently close for the profit margins I outlined.

As far as 250W, I just googled power use of a system with a 6870. I'd have to get a kill-a-watt or ups to get a better number.

I figured the numbers I had been using would be accurate enough for me to do some math to figure out feasibility.
 
That is extremely expensive electricity. For you, it works out to $5.2 in Electricity for 1 BTC ($6-$7 right now). With luck you will make $1. With bad luck, you will lose money.

I pay $0.07/ KWh for electricity. I have approximately 4200MH/s across all miners using 2500W = <$5 in electricity per day. Selling @ $7 means $21/day = ~$16/day profit

Turned all these numbers into an excel spreadsheet (any interest anyone?) and plugged in you numbers at my electricity rate and I'd make $11.28 a day. Of course I could probably add another $50 I save on my heating bill, lol. That is all at the current rate of $6.28 BTC.
 
Turned all these numbers into an excel spreadsheet (any interest anyone?) and plugged in you numbers at my electricity rate and I'd make $11.28 a day. Of course I could probably add another $50 I save on my heating bill, lol. That is all at the current rate of $6.28 BTC.

Google docs, with public viewing?

Anyways, for those of us who are looking into upgrading and beginning mining, the difficulty is going to double this year!
 
Google docs, with public viewing?

Anyways, for those of us who are looking into upgrading and beginning mining, the difficulty is going to double this year!

not precisely accurate

the reward per block found will halve, true, this will likely push marginal operators out and the difficulty will likely drop somewhat

you are correct that the net overall effect will likely be lower BTC/Watt payout to miners
 
Last edited:
Has anyone tried running both AMD and nvidia cards in the same box? I'm thinking about running a used AMD in my folding box to earn some cash and don't want to risk too much by building a new rig specifically for BTC in case it tanks. I have access to free electric.
 
Google docs, with public viewing?

Anyways, for those of us who are looking into upgrading and beginning mining, the difficulty is going to double this year!

I can do that on Monday. I have the excel file on my work computer.
 
Back
Top