Author

Topic: Vanitygen: Vanity bitcoin address generator/miner [v0.22] - page 152. (Read 1153383 times)

legendary
Activity: 1792
Merit: 1008
/dev/null
I just found this, and I must say, it's awesome! Getting ~30Mkey/s on my 7970 @ 1150/1485.
Quote
on 7970:
(seems it is finding matches, but disregarding due to mismatch?
Code:
CPU hash: 6707a76e848f5e9368c2d9d9ec6d60880df44891
GPU hash: 020b0bda39a9bbb2eeddd194a8be20934835dff6
Found delta: 94027 Start delta: 1
When you get the CPU hash/GPU hash mismatch, it means the OpenCL device is producing incorrect results.  I have not had a chance to test it on an AMD 7XXX series GCN card yet, but others have reported that it works.  Try running it with -V on that card for a little while and see what you get.  Since it's producing incorrect results, the performance numbers may or may not mean anything.
Looks like I spoke too soon, cuz I'm actually getting this this error on my 7970, with 12.8 drivers. I tried a few different flag options, but nothing really worked.
Well I only get 9Mkey/s instead of 30, but the -S option produces valid results.
Still not working with the newer 12.10 drivers. Is this a problem with the program? If oclvanitygen produces invalid results, i'm assuming oclvanityminer will be the same?
probably just usual ATI crap driver, hopefully it will get fixed or there will be a workaround.
But I've heard of other using 5xxx cards, and it works fine. This just seems to be an issue with the 7xxx cards. Like I said, the -S option works fine, but at 1/3 of the speed.
and thats what ive said, the OpenCL driver is broken for 7*** cards.
legendary
Activity: 952
Merit: 1000
I just found this, and I must say, it's awesome! Getting ~30Mkey/s on my 7970 @ 1150/1485.
Quote
on 7970:
(seems it is finding matches, but disregarding due to mismatch?
Code:
CPU hash: 6707a76e848f5e9368c2d9d9ec6d60880df44891
GPU hash: 020b0bda39a9bbb2eeddd194a8be20934835dff6
Found delta: 94027 Start delta: 1
When you get the CPU hash/GPU hash mismatch, it means the OpenCL device is producing incorrect results.  I have not had a chance to test it on an AMD 7XXX series GCN card yet, but others have reported that it works.  Try running it with -V on that card for a little while and see what you get.  Since it's producing incorrect results, the performance numbers may or may not mean anything.
Looks like I spoke too soon, cuz I'm actually getting this this error on my 7970, with 12.8 drivers. I tried a few different flag options, but nothing really worked.
Well I only get 9Mkey/s instead of 30, but the -S option produces valid results.
Still not working with the newer 12.10 drivers. Is this a problem with the program? If oclvanitygen produces invalid results, i'm assuming oclvanityminer will be the same?
probably just usual ATI crap driver, hopefully it will get fixed or there will be a workaround.
But I've heard of other using 5xxx cards, and it works fine. This just seems to be an issue with the 7xxx cards. Like I said, the -S option works fine, but at 1/3 of the speed.
legendary
Activity: 1792
Merit: 1008
/dev/null
I just found this, and I must say, it's awesome! Getting ~30Mkey/s on my 7970 @ 1150/1485.
Quote
on 7970:
(seems it is finding matches, but disregarding due to mismatch?
Code:
CPU hash: 6707a76e848f5e9368c2d9d9ec6d60880df44891
GPU hash: 020b0bda39a9bbb2eeddd194a8be20934835dff6
Found delta: 94027 Start delta: 1
When you get the CPU hash/GPU hash mismatch, it means the OpenCL device is producing incorrect results.  I have not had a chance to test it on an AMD 7XXX series GCN card yet, but others have reported that it works.  Try running it with -V on that card for a little while and see what you get.  Since it's producing incorrect results, the performance numbers may or may not mean anything.
Looks like I spoke too soon, cuz I'm actually getting this this error on my 7970, with 12.8 drivers. I tried a few different flag options, but nothing really worked.
Well I only get 9Mkey/s instead of 30, but the -S option produces valid results.
Still not working with the newer 12.10 drivers. Is this a problem with the program? If oclvanitygen produces invalid results, i'm assuming oclvanityminer will be the same?
probably just usual ATI crap driver, hopefully it will get fixed or there will be a workaround.
legendary
Activity: 952
Merit: 1000
I just found this, and I must say, it's awesome! Getting ~30Mkey/s on my 7970 @ 1150/1485.
Quote
on 7970:
(seems it is finding matches, but disregarding due to mismatch?
Code:
CPU hash: 6707a76e848f5e9368c2d9d9ec6d60880df44891
GPU hash: 020b0bda39a9bbb2eeddd194a8be20934835dff6
Found delta: 94027 Start delta: 1
When you get the CPU hash/GPU hash mismatch, it means the OpenCL device is producing incorrect results.  I have not had a chance to test it on an AMD 7XXX series GCN card yet, but others have reported that it works.  Try running it with -V on that card for a little while and see what you get.  Since it's producing incorrect results, the performance numbers may or may not mean anything.
Looks like I spoke too soon, cuz I'm actually getting this this error on my 7970, with 12.8 drivers. I tried a few different flag options, but nothing really worked.
Well I only get 9Mkey/s instead of 30, but the -S option produces valid results.
Still not working with the newer 12.10 drivers. Is this a problem with the program? If oclvanitygen produces invalid results, i'm assuming oclvanityminer will be the same?
legendary
Activity: 1792
Merit: 1008
/dev/null
i get >20 on a 5830 running at 980/1025 1280x1024

I find it strange that I get 25 Mkey/s on Windows 7, but only 20 Mkey/s on linux. I tried upgrading to the bleeding edge xorg PPA and that made no difference.

I neglected to mention that these two figures are from the SAME MACHINE - dual booted between windows 7 and Ubuntu 12.04. So, same video card, same motherboard, etc.
thats simple, ati drivers for linux are same as winblows; crap, on linux even alot more!
try upgrade to the latest fglrx and picking the right SDK, good luck Wink
Also, what versions of openssl?

On linux:
libcurl4-openssl-dev                   7.22.0-3ubuntu4
openssl                                1.0.1-4ubuntu5.5

On windows - I think the executable is statically linked with SSL? It's the latest version of oclvanityminer64

libcurl4-openssl-dev
hero member
Activity: 481
Merit: 500
i get >20 on a 5830 running at 980/1025 1280x1024

I find it strange that I get 25 Mkey/s on Windows 7, but only 20 Mkey/s on linux. I tried upgrading to the bleeding edge xorg PPA and that made no difference.

I neglected to mention that these two figures are from the SAME MACHINE - dual booted between windows 7 and Ubuntu 12.04. So, same video card, same motherboard, etc.
thats simple, ati drivers for linux are same as winblows; crap, on linux even alot more!
try upgrade to the latest fglrx and picking the right SDK, good luck Wink
Also, what versions of openssl?

On linux:
libcurl4-openssl-dev                   7.22.0-3ubuntu4
openssl                                1.0.1-4ubuntu5.5

On windows - I think the executable is statically linked with SSL? It's the latest version of oclvanityminer64
hero member
Activity: 742
Merit: 500
i get >20 on a 5830 running at 980/1025 1280x1024

I find it strange that I get 25 Mkey/s on Windows 7, but only 20 Mkey/s on linux. I tried upgrading to the bleeding edge xorg PPA and that made no difference.

I neglected to mention that these two figures are from the SAME MACHINE - dual booted between windows 7 and Ubuntu 12.04. So, same video card, same motherboard, etc.
thats simple, ati drivers for linux are same as winblows; crap, on linux even alot more!
try upgrade to the latest fglrx and picking the right SDK, good luck Wink
Also, what versions of openssl?
legendary
Activity: 1792
Merit: 1008
/dev/null
i get >20 on a 5830 running at 980/1025 1280x1024

I find it strange that I get 25 Mkey/s on Windows 7, but only 20 Mkey/s on linux. I tried upgrading to the bleeding edge xorg PPA and that made no difference.

I neglected to mention that these two figures are from the SAME MACHINE - dual booted between windows 7 and Ubuntu 12.04. So, same video card, same motherboard, etc.
thats simple, ati drivers for linux are same as winblows; crap, on linux even alot more!
try upgrade to the latest fglrx and picking the right SDK, good luck Wink
hero member
Activity: 481
Merit: 500
i get >20 on a 5830 running at 980/1025 1280x1024

I find it strange that I get 25 Mkey/s on Windows 7, but only 20 Mkey/s on linux. I tried upgrading to the bleeding edge xorg PPA and that made no difference.

I neglected to mention that these two figures are from the SAME MACHINE - dual booted between windows 7 and Ubuntu 12.04. So, same video card, same motherboard, etc.
full member
Activity: 120
Merit: 100
Code:
C:\Users\...\vanitygen-0.22-win>oclvanityminer -u https://vanitypool.appspot.com/ -a 1N#########################k
Get work request failed: SSL connect error

Oclvanitygen uses libcurl, and unfortunately it won't read your proxy settings for MSIE or Firefox.  If you need to set a proxy, you can set the http_proxy or HTTPS_PROXY environment variables.  For example:

Code:
H:\>set HTTPS_PROXY="http://172.16.0.1:8080/"
H:\>oclvanityminer -u https://vanitypool.appspot.com -a 1XXXXXXXXXXXXXXXXXXXX
thanks worked perfectly.
Is it possible to run two devices at once or can I only do this by opening oclvanityminer twice?
and how can I select a particular pattern to mine?
hero member
Activity: 720
Merit: 525
I finally had a chance to look at the source and apparently the units of the "value" shown are simply BTC/Mkey, but multiplied by 3600, or in code:

Code:
wip->value = (reward * 1000000.0 * 3600.0) / difficulty;

This is not BTC/MKeyHr, this is a meaningless number. More correctly, you could say the units are BTC*s/Mkey*hr. I suggest removing the 3600 and just displaying the value as BTC/Mkey. If the number is too small (too many decimals to be easily readable), change it to BTC/Gkey, since most people are calculating at least a Gkey every minute. If you would like to go one step further, you can show BTC/hr as:

Code:
3600 * rate * reward / difficulty

Where rate is the calculated Mkey/s for your hardware (so vg_output_timing() will have to been called already).

Hi fizzisist,

You're right, it's an entirely arbitrary and not 100% correctly labelled unit.  I just wanted it to display some semi-comparable notion of value.  The idea was to create something along the lines of kilowatt-hours, whereas the proper unit would be megajoules, but the power company doesn't publish their electricity rates per megajoule.

Anyway, I'll take your patch on GitHub to convert the display unit to GKeys.  Thanks!

Great! I completely understood where you were coming from once I switched it to BTC/Mkey and all the values displayed as 0.000000! BTC/Gkey is definitely better than BTC/Mkey. BTC/MkeyHr isn't that confusing (bitcoindaddy figured it out above), but it's unclear enough.

I'm more happy to see the "Total value" line, especially when there are a lot of patterns available from a single public key, like there are right now! Thanks for pulling that in!
legendary
Activity: 1792
Merit: 1008
/dev/null
i get >20 on a 5830 running at 980/1025 1280x1024

I find it strange that I get 25 Mkey/s on Windows 7, but only 20 Mkey/s on linux. I tried upgrading to the bleeding edge xorg PPA and that made no difference.
use bfgminer -> set intensity to d -> enjoy

Sorry, what does bfgminer have to do with oclvanityminer?
sry wrong thread Tongue
legendary
Activity: 916
Merit: 1003
Sorry, what does bfgminer have to do with oclvanityminer?

Absolutely nothing.
hero member
Activity: 481
Merit: 500
i get >20 on a 5830 running at 980/1025 1280x1024

I find it strange that I get 25 Mkey/s on Windows 7, but only 20 Mkey/s on linux. I tried upgrading to the bleeding edge xorg PPA and that made no difference.
use bfgminer -> set intensity to d -> enjoy

Sorry, what does bfgminer have to do with oclvanityminer?
legendary
Activity: 1792
Merit: 1008
/dev/null
i get >20 on a 5830 running at 980/1025 1280x1024

I find it strange that I get 25 Mkey/s on Windows 7, but only 20 Mkey/s on linux. I tried upgrading to the bleeding edge xorg PPA and that made no difference.
use bfgminer -> set intensity to d -> enjoy
hero member
Activity: 481
Merit: 500
i get >20 on a 5830 running at 980/1025 1280x1024

I find it strange that I get 25 Mkey/s on Windows 7, but only 20 Mkey/s on linux. I tried upgrading to the bleeding edge xorg PPA and that made no difference.
zvs
legendary
Activity: 1680
Merit: 1000
https://web.archive.org/web/*/nogleg.com
Please help me understand this:

Just started experimenting with samr7/oclvanityminer and have dedicated a 5850 to the task. According to the calculator, I'll earn a little over .11BTC per day with this card doing normal (non-vanity) mining.

It's currently hashing around 20.20 Mkey/s and working on this problem:

"1vanity" Reward: 0.079600 Value: 0.000323 BTC/MkeyHr.

So that's 0.000323 * 20.20 * 24 = 0.1565904 BTC, right? So I should be earning more than I would if I mined straight? But then they take 20 percent out first or does that figure already have the 20 percent taken out?

It also says 50% in 8.0h, I take that to mean there's a 50 percent chance it will find the solution in 8 hours. So approximately 100% (statistically) in 16 hours, that means the reward (0.796) for 16 hours average work?  So 24/16 * 0.0796 BTC = 0.1194 BTC which is not close to 0.1565904 BTC. Where is my math failing?

Does my Mkey/s look reasonable for that card?

i get >20 on a 5830 running at 980/1025 1280x1024
full member
Activity: 140
Merit: 430
Firstbits: 1samr7
I finally had a chance to look at the source and apparently the units of the "value" shown are simply BTC/Mkey, but multiplied by 3600, or in code:

Code:
wip->value = (reward * 1000000.0 * 3600.0) / difficulty;

This is not BTC/MKeyHr, this is a meaningless number. More correctly, you could say the units are BTC*s/Mkey*hr. I suggest removing the 3600 and just displaying the value as BTC/Mkey. If the number is too small (too many decimals to be easily readable), change it to BTC/Gkey, since most people are calculating at least a Gkey every minute. If you would like to go one step further, you can show BTC/hr as:

Code:
3600 * rate * reward / difficulty

Where rate is the calculated Mkey/s for your hardware (so vg_output_timing() will have to been called already).

Hi fizzisist,

You're right, it's an entirely arbitrary and not 100% correctly labelled unit.  I just wanted it to display some semi-comparable notion of value.  The idea was to create something along the lines of kilowatt-hours, whereas the proper unit would be megajoules, but the power company doesn't publish their electricity rates per megajoule.

Anyway, I'll take your patch on GitHub to convert the display unit to GKeys.  Thanks!
full member
Activity: 140
Merit: 430
Firstbits: 1samr7
Code:
C:\Users\...\vanitygen-0.22-win>oclvanityminer -u https://vanitypool.appspot.com/ -a 1N#########################k
Get work request failed: SSL connect error

Been trying the same thing for an hour, any other suggestions? I'd really appreciate it.

Hi wigglyuk,

Do you use an HTTP proxy?

Oclvanitygen uses libcurl, and unfortunately it won't read your proxy settings for MSIE or Firefox.  If you need to set a proxy, you can set the http_proxy or HTTPS_PROXY environment variables.  For example:

Code:
H:\>set HTTPS_PROXY="http://172.16.0.1:8080/"
H:\>oclvanityminer -u https://vanitypool.appspot.com -a 1XXXXXXXXXXXXXXXXXXXX
hero member
Activity: 720
Merit: 525
Please help me understand this:

Just started experimenting with samr7/oclvanityminer and have dedicated a 5850 to the task. According to the calculator, I'll earn a little over .11BTC per day with this card doing normal (non-vanity) mining.

It's currently hashing around 20.20 Mkey/s and working on this problem:

"1vanity" Reward: 0.079600 Value: 0.000323 BTC/MkeyHr.

So that's 0.000323 * 20.20 * 24 = 0.1565904 BTC, right? So I should be earning more than I would if I mined straight? But then they take 20 percent out first or does that figure already have the 20 percent taken out?

It also says 50% in 8.0h, I take that to mean there's a 50 percent chance it will find the solution in 8 hours. So approximately 100% (statistically) in 16 hours, that means the reward (0.796) for 16 hours average work?  So 24/16 * 0.0796 BTC = 0.1194 BTC which is not close to 0.1565904 BTC. Where is my math failing?

Does my Mkey/s look reasonable for that card?

I was interested in this too and realized there was very little out there to help people figure this out, so I wrote some scripts to do the calculations for me: http://fizzisist.com/mining-value

As for the statistical stuff, the BTC/Mkey stuff is the number you should be using. This is the expected earnings on average with the current pool of work. It doesn't matter how long you mine for (although as you mine for longer time, your earnings will approach that of the average, yet can be subject to huge statistical variance over short time periods). Bitcoin solo mining has the same problem, so pools were invented. Unfortunately, vanity pool isn't a pool in the same sense, so you still have quite a bit of variance, although MUCH less than solo mining bitcoins.
Jump to: