bigjme
|
|
April 29, 2014, 12:38:34 PM Last edit: April 29, 2014, 12:51:17 PM by bigjme |
|
doesn't look like my fix works, or even compiles, i knew it was the algo issue. it seems like it just doesnt set the algorithm up properly. have you tested it mining to make sure it changes the algo from heavycoin?
the config selection is soooo messy its unreal me thinks it needs re-writing
|
Owner of: cudamining.co.uk
|
|
|
|
|
|
|
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
|
ltcnim
Legendary
Offline
Activity: 914
Merit: 1001
|
|
April 29, 2014, 12:59:46 PM |
|
alright, I'm kinda confused right now. I haven't mined groestl in a while (because of the low profit), but I wanted to give the new ccminer a try. So I compiled it from github (after fixing the spelling error in both .cu files), and the result for my 5x750Ti rig is not what I expected. I get 13000khash for the rig, which means it's 2600khash per card. I'm running ubuntu 13.10 so, obviously, no overclocks.
The strange thing is, that my profitability calculator which I wrote a while back (and which basicly decides what to mine) has 16000khash in it. And I'm pretty much sure, that I got that khash when I last mined groestl. Also, there is absolutely no other coin which I could have confused with groestl.
I will try the old 0.5 version again, but I wonder if the old 16000khash value is possible for my stock clock rig?
|
|
|
|
bigjme
|
|
April 29, 2014, 01:01:05 PM |
|
alright, I'm kinda confused right now. I haven't mined groestl in a while (because of the low profit), but I wanted to give the new ccminer a try. So I compiled it from github (after fixing the spelling error in both .cu files), and the result for my 5x750Ti rig is not what I expected. I get 13000khash for the rig, which means it's 2600khash per card. I'm running ubuntu 13.10 so, obviously, no overclocks.
The strange thing is, that my profitability calculator which I wrote a while back (and which basicly decides what to mine) has 16000khash in it. And I'm pretty much sure, that I got that khash when I last mined groestl. Also, there is absolutely no other coin which I could have confused with groestl.
I will try the old 0.5 version again, but I wonder if the old 16000khash value is possible for my stock clock rig?
you fixed the spelling error but did you set #define MAXWELL_OR_FERMI 1 in both files?
|
Owner of: cudamining.co.uk
|
|
|
ltcnim
Legendary
Offline
Activity: 914
Merit: 1001
|
|
April 29, 2014, 01:04:32 PM |
|
uhm...well no. I thought it's either change the spelling error "MAXWELL_OR_FEMRI" OR change the definition? But I might have missed that part. I'll give it a try, thanks!
|
|
|
|
AizenSou
|
|
April 29, 2014, 01:06:25 PM |
|
doesn't look like my fix works, or even compiles, i knew it was the algo issue. it seems like it just doesnt set the algorithm up properly. have you tested it mining to make sure it changes the algo from heavycoin?
the config selection is soooo messy its unreal me thinks it needs re-writing
Here is my fix: quick & dirty EDIT: I managed to solve the bugs by deleting these two if clause in parse_config and parse_cmdline function. Ugly quick fix but it works: if (opt_algo == ALGO_HEAVY && opt_vote == 9999) { fprintf(stderr, "%s: Heavycoin hash requires block reward vote parameter (see --vote)\n", argv[0]); show_usage_and_exit(1); }
|
|
|
|
bigjme
|
|
April 29, 2014, 01:11:23 PM |
|
Here is my fix: quick & dirty EDIT: I managed to solve the bugs by deleting these two if clause in parse_config and parse_cmdline function. Ugly quick fix but it works: if (opt_algo == ALGO_HEAVY && opt_vote == 9999) { fprintf(stderr, "%s: Heavycoin hash requires block reward vote parameter (see --vote)\n", argv[0]); show_usage_and_exit(1); } yeh i found that, it does select the correct algo right?
|
Owner of: cudamining.co.uk
|
|
|
ltcnim
Legendary
Offline
Activity: 914
Merit: 1001
|
|
April 29, 2014, 01:16:57 PM |
|
alright, I'm kinda confused right now. I haven't mined groestl in a while (because of the low profit), but I wanted to give the new ccminer a try. So I compiled it from github (after fixing the spelling error in both .cu files), and the result for my 5x750Ti rig is not what I expected. I get 13000khash for the rig, which means it's 2600khash per card. I'm running ubuntu 13.10 so, obviously, no overclocks.
The strange thing is, that my profitability calculator which I wrote a while back (and which basicly decides what to mine) has 16000khash in it. And I'm pretty much sure, that I got that khash when I last mined groestl. Also, there is absolutely no other coin which I could have confused with groestl.
I will try the old 0.5 version again, but I wonder if the old 16000khash value is possible for my stock clock rig?
you fixed the spelling error but did you set #define MAXWELL_OR_FERMI 1 in both files? ok, that was indeed the problem, now I'm back @ 16000khash. Thanks. Unfortunately, the profitability is way too low (0.00315/day) which means, that even if the algo can be boosted to 200% of the current speed, it's not profitable compared to other coins
|
|
|
|
bigjme
|
|
April 29, 2014, 01:21:59 PM |
|
ok, that was indeed the problem, now I'm back @ 16000khash. Thanks. Unfortunately, the profitability is way too low (0.00315/day) which means, that even if the algo can be boosted to 200% of the current speed, it's not profitable compared to other coins i am earning 0.0048312 BTC a day on 3 750Ti's after taking out power usage of my entire room, that is including my dac and audio stuff, i am still maing £0.67 a day the improvement is said to triple performance, so that would boost my daily income to 0.0144936 BTC or or an income of £118 a month, enough to pay for an entire card every month, even if i cover the £17 power bill out my own pocket
|
Owner of: cudamining.co.uk
|
|
|
ltcnim
Legendary
Offline
Activity: 914
Merit: 1001
|
|
April 29, 2014, 01:25:46 PM |
|
ok, that was indeed the problem, now I'm back @ 16000khash. Thanks. Unfortunately, the profitability is way too low (0.00315/day) which means, that even if the algo can be boosted to 200% of the current speed, it's not profitable compared to other coins i am earning 0.0048312 BTC a day on 3 750Ti's after taking out power usage of my entire room, that is including my dac and audio stuff, i am still maing £0.67 a day the improvement is said to triple performance, so that would boost my daily income to 0.0144936 BTC or or an income of £118 a month, enough to pay for an entire card every month, even if i cover the £17 power bill out my own pocket what is your calculation for that? I take realtime values for each calculation (diff/price). At the moment I write it's: diff: 199.4977 price: 0.00000580 speed: 16000khash/s which results in: 0.0033052247 BTC a day
|
|
|
|
bigjme
|
|
April 29, 2014, 01:29:48 PM |
|
i am calculating from what i am earning in 24 hours, currently 730grs, mintpals current avg price is 0.00000650, thats at 10.5MH/s
730 * 0.00000650 = 0.004745BTC
can you tell me how you calculated the stuff from diff etc. i really want to do a proper realtime profit calc
|
Owner of: cudamining.co.uk
|
|
|
AizenSou
|
|
April 29, 2014, 01:34:59 PM |
|
i am calculating from what i am earning in 24 hours, currently 730grs, mintpals current avg price is 0.00000650, thats at 10.5MH/s
730 * 0.00000650 = 0.004745BTC
can you tell me how you calculated the stuff from diff etc. i really want to do a proper realtime profit calc
Great article for this info with excellent explanation: http://www.holynerdvana.com/2014/02/how-to-calculate-coins-per-day-for-any.htmlBtw don't mine groestl if you want profit. Mine scrypt-jane (N=>12). My tips for today (please don't tell Christian ): mine Jackpotcoin. My 30x750Tis farm earned me 100$ for last 30 hours. Greets,
|
|
|
|
cbuchner1 (OP)
|
|
April 29, 2014, 01:35:53 PM |
|
Lmao. I can not wait to get the new groestl code. Its going to be amazing! Are you still waiting to get x11 finished before you release it christian?
two more hash functions need bugfixing and polish, then X11 will run adequately. The SIMD hash is particularly nasty. can we know what's the estimated hashrate for x11 750ti ? "adequate" adequate means one 750 Ti beats one R9 290. That's what we're aiming for. Christian
|
|
|
|
bigjme
|
|
April 29, 2014, 01:39:30 PM |
|
is jackpotcoin even selling? luck with 3 750Ti's solo mining may not be so good
|
Owner of: cudamining.co.uk
|
|
|
cbuchner1 (OP)
|
|
April 29, 2014, 01:39:42 PM |
|
@christian
maybe you can understand this in ccminer even if i set compute_20 and sm_20 it still compiles using -gencode=arch=compute_35,code=\"sm_35,compute_35\" now i know sm_20 is literally a shortcut for -arch=compute_20
is it passing it from somewhere else other then the project=>cuda c/c++=>device?
on Linux we've set it up so all compute versions end up in the same binary. I wish it was easier to do the same for Windows. The only thing that is really missing is an automatic code path selection in the table based Groestl code (that's the ugly #define MAXWELL_OR_FERMI) Christian
|
|
|
|
ltcnim
Legendary
Offline
Activity: 914
Merit: 1001
|
|
April 29, 2014, 01:45:03 PM |
|
i am calculating from what i am earning in 24 hours, currently 730grs, mintpals current avg price is 0.00000650, thats at 10.5MH/s
730 * 0.00000650 = 0.004745BTC
can you tell me how you calculated the stuff from diff etc. i really want to do a proper realtime profit calc
basicly, it's: (1 / (86400 / (difficulty * 2^32 / hashrate))) * block reward to get the coins you earn each day (where 86400 is the number of seconds in 24h). From there, the rest is simple. Coins * current price. A few things to notice: 1. difficulty jumps like crazy, so groestl for example is 199.54 now, and the next seconds it jumps to 263 which of course has a big impact on profit. Due to the fast re-allocation times of difficulty you would need to get an average value over lets say the last hour to get better results. Also, the price is a bit tricky. Sometimes the last price you get from the exchanges is only a fake buy of one coin where the average price is much lower.
|
|
|
|
cayars
|
|
April 29, 2014, 01:45:09 PM |
|
Asked before but never got an answer.
Will ccminer and cudaminer ever be merged or are they now on definite separate paths?
|
|
|
|
cayars
|
|
April 29, 2014, 01:47:43 PM |
|
i am calculating from what i am earning in 24 hours, currently 730grs, mintpals current avg price is 0.00000650, thats at 10.5MH/s
730 * 0.00000650 = 0.004745BTC
can you tell me how you calculated the stuff from diff etc. i really want to do a proper realtime profit calc
Great article for this info with excellent explanation: http://www.holynerdvana.com/2014/02/how-to-calculate-coins-per-day-for-any.htmlBtw don't mine groestl if you want profit. Mine scrypt-jane (N=>12). My tips for today (please don't tell Christian ): mine Jackpotcoin. My 30x750Tis farm earned me 100$ for last 30 hours. Greets, Check YAC and see if that would be more profitable for you.
|
|
|
|
AizenSou
|
|
April 29, 2014, 01:49:27 PM |
|
i am calculating from what i am earning in 24 hours, currently 730grs, mintpals current avg price is 0.00000650, thats at 10.5MH/s
730 * 0.00000650 = 0.004745BTC
can you tell me how you calculated the stuff from diff etc. i really want to do a proper realtime profit calc
Great article for this info with excellent explanation: http://www.holynerdvana.com/2014/02/how-to-calculate-coins-per-day-for-any.htmlBtw don't mine groestl if you want profit. Mine scrypt-jane (N=>12). My tips for today (please don't tell Christian ): mine Jackpotcoin. My 30x750Tis farm earned me 100$ for last 30 hours. Greets, Check YAC and see if that would be more profitable for you. This is what I mean with scrypt-jane N>=12 coins . YAC includes.
|
|
|
|
bigjme
|
|
April 29, 2014, 01:50:20 PM |
|
weirdly for jackpot coin i have a stake of over 800000 but no balance, even with over 3000 confirms grr its so annoying, i use to mine yac etc. and now i cant even remember
i need to find a launch config for my 750Ti's then find somewhere to mine at, last time i tried i got nothing but issues
|
Owner of: cudamining.co.uk
|
|
|
cayars
|
|
April 29, 2014, 01:56:56 PM |
|
i am calculating from what i am earning in 24 hours, currently 730grs, mintpals current avg price is 0.00000650, thats at 10.5MH/s
730 * 0.00000650 = 0.004745BTC
can you tell me how you calculated the stuff from diff etc. i really want to do a proper realtime profit calc
Great article for this info with excellent explanation: http://www.holynerdvana.com/2014/02/how-to-calculate-coins-per-day-for-any.htmlBtw don't mine groestl if you want profit. Mine scrypt-jane (N=>12). My tips for today (please don't tell Christian ): mine Jackpotcoin. My 30x750Tis farm earned me 100$ for last 30 hours. Greets, Check YAC and see if that would be more profitable for you. This is what I mean with scrypt-jane N>=12 coins . YAC includes. I should have been more specific. IE Would you make more RIGHT NOW mining YAC instead of Jackpotcoin?
|
|
|
|
|