Bitcoin Forum
December 10, 2016, 09:10:55 AM *
News: Latest stable version of Bitcoin Core: 0.13.1  [Torrent].
 
   Home   Help Search Donate Login Register  
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 [30] 31 »
  Print  
Author Topic: ATTN Litecoin GPU Miners - Scrypt support for cgminer  (Read 169334 times)
Luke-Jr
Legendary
*
Offline Offline

Activity: 2100



View Profile
August 01, 2012, 07:33:02 AM
 #581

so if i am not mistaken the main difference between LTC and BTC was the fact that LTC was targeted at CPU mining only.

now that you can mine it on GPU, arent both currencies basically the same? making one redundant?
Litecoin was always redundant, and a scam. There is reason to believe the original designer of the scrypt algorithm had been GPU mining it from the first day. Also, it's impossible to make a CPU-only proof-of-work. Ironically, now that GPU mining is public, they're claiming it's ASIC-resistant which is even more absurd (nothing can be ASIC resistant).

1481361055
Hero Member
*
Offline Offline

Posts: 1481361055

View Profile Personal Message (Offline)

Ignore
1481361055
Reply with quote  #2

1481361055
Report to moderator
1481361055
Hero Member
*
Offline Offline

Posts: 1481361055

View Profile Personal Message (Offline)

Ignore
1481361055
Reply with quote  #2

1481361055
Report to moderator
1481361055
Hero Member
*
Offline Offline

Posts: 1481361055

View Profile Personal Message (Offline)

Ignore
1481361055
Reply with quote  #2

1481361055
Report to moderator
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1481361055
Hero Member
*
Offline Offline

Posts: 1481361055

View Profile Personal Message (Offline)

Ignore
1481361055
Reply with quote  #2

1481361055
Report to moderator
Tittiez
Hero Member
*****
Offline Offline

Activity: 686



View Profile
August 01, 2012, 07:54:12 AM
 #582

There is already a windows build from ckolivas, why would someone download a windows version from somewhere else?

At the time of writing it there wasn't.
smoothie
Legendary
*
Offline Offline

Activity: 1848


LEALANA Monero Physical Silver Coins


View Profile
August 01, 2012, 08:45:49 AM
 #583

so if i am not mistaken the main difference between LTC and BTC was the fact that LTC was targeted at CPU mining only.

now that you can mine it on GPU, arent both currencies basically the same? making one redundant?
Litecoin was always redundant, and a scam. There is reason to believe the original designer of the scrypt algorithm had been GPU mining it from the first day. Also, it's impossible to make a CPU-only proof-of-work. Ironically, now that GPU mining is public, they're claiming it's ASIC-resistant which is even more absurd (nothing can be ASIC resistant).

1. How is it a scam? Draw me a picture please.

2. Who lost money?

3. How is price fluctuations (like bitcoin) make it any more or less of a "scam"?

4. Wasn't ArtForz mining bitcoins with his GPU miner before any publicly released GPU miner existed? How is that any different (if it is true, which you have no proof of)?

5. "Nothing can be ASIC resistant". Well what you are doing is taking statements that are made by people who dont have a clue how litecoin works and using that to portray it as if every litecoin support believes that.

Do us all a set of favors:

1. Stop being a Douche.

2. Stop spreading bullshit and claiming that all LTC supporters  agree with certain statements.

3. Grow the fuck up. Most of the other early adopters/devs aren't as fucking childish as you are.

My rant has ended. Have a nice day!

 Cheesy Cheesy Cheesy

███████████████████████████████████████

            ,╓p@@███████@╗╖,           
        ,p████████████████████N,       
      d█████████████████████████b     
    d██████████████████████████████æ   
  ,████²█████████████████████████████, 
 ,█████  ╙████████████████████╨  █████y
 ██████    `████████████████`    ██████
║██████       Ñ███████████`      ███████
███████         ╩██████Ñ         ███████
███████    ▐▄     ²██╩     a▌    ███████
╢██████    ▐▓█▄          ▄█▓▌    ███████
 ██████    ▐▓▓▓▓▌,     ▄█▓▓▓▌    ██████─
           ▐▓▓▓▓▓▓█,,▄▓▓▓▓▓▓▌          
           ▐▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▌          
    ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓─  
     ²▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓╩    
        ▀▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▀       
           ²▀▀▓▓▓▓▓▓▓▓▓▓▓▓▀▀`          
                   ²²²                 
███████████████████████████████████████

. ★☆ WWW.LEALANA.COM        My PGP fingerprint is A764D833.        SMOOTHIE'S HEALTH AND FITNESS JOURNAL          History of Monero development Visualization ★☆ .
LEALANA  PHYSICAL MONERO COINS 999 FINE SILVER.
 
bitlane
Internet detective
Sr. Member
****
Offline Offline

Activity: 462


I heart thebaron


View Profile
August 01, 2012, 08:50:35 AM
 #584

I seem to be experiencing a problem with the new CGMiner 2.6.1-official that was not affecting 2.5.0.Scrypt_beta_ee

CGMiner now seems to be seeing cards at times as sick and the program crashes. It has happened to each of my miners at least once since I switched from the 'EE' Beta 2.5.0 yesterday, to 2.6.1 Official.

The last beta ('EE') seemed to run fine for multiple days, but I have yet to experience a SOLID 24 hours from 2.6.1-Official.....without a crash.

The 'EE' version, for reference...
win32: http://www.mediafire.com/?0yxa1cf35pk5q2i

Latest Commits included!

-ck
Staff
Legendary
*
Offline Offline

Activity: 2002


Ruu \o/


View Profile WWW
August 01, 2012, 08:54:38 AM
 #585

so if i am not mistaken the main difference between LTC and BTC was the fact that LTC was targeted at CPU mining only.

now that you can mine it on GPU, arent both currencies basically the same? making one redundant?
Litecoin was always redundant, and a scam. There is reason to believe the original designer of the scrypt algorithm had been GPU mining it from the first day. Also, it's impossible to make a CPU-only proof-of-work. Ironically, now that GPU mining is public, they're claiming it's ASIC-resistant which is even more absurd (nothing can be ASIC resistant).

1. How is it a scam? Draw me a picture please.

2. Who lost money?

3. How is price fluctuations (like bitcoin) make it any more or less of a "scam"?

4. Wasn't ArtForz mining bitcoins with his GPU miner before any publicly released GPU miner existed? How is that any different (if it is true, which you have no proof of)?

5. "Nothing can be ASIC resistant". Well what you are doing is taking statements that are made by people who dont have a clue how litecoin works and using that to portray it as if every litecoin support believes that.

Do us all a set of favors:

1. Stop being a Douche.

2. Stop spreading bullshit and claiming that all LTC supporters  agree with certain statements.

3. Grow the fuck up. Most of the other early adopters/devs aren't as fucking childish as you are.

My rant has ended. Have a nice day!

 Cheesy Cheesy Cheesy
Fatal mistake. Don't respond to trolls. The one thing that is MOST effective against them is ignoring them, or at most, talking about them in the 3rd person as though they're not present.

Luke-jr has failed to accept one of the universal truths: If you end up having a problem with the whole world, then perhaps it's not the whole world at fault.

Primary developer/maintainer for cgminer and ckpool/ckproxy.
Pooled mine at kano.is, solo mine at solo.ckpool.org
-ck
smoothie
Legendary
*
Offline Offline

Activity: 1848


LEALANA Monero Physical Silver Coins


View Profile
August 01, 2012, 09:01:56 AM
 #586

so if i am not mistaken the main difference between LTC and BTC was the fact that LTC was targeted at CPU mining only.

now that you can mine it on GPU, arent both currencies basically the same? making one redundant?
Litecoin was always redundant, and a scam. There is reason to believe the original designer of the scrypt algorithm had been GPU mining it from the first day. Also, it's impossible to make a CPU-only proof-of-work. Ironically, now that GPU mining is public, they're claiming it's ASIC-resistant which is even more absurd (nothing can be ASIC resistant).

1. How is it a scam? Draw me a picture please.

2. Who lost money?

3. How is price fluctuations (like bitcoin) make it any more or less of a "scam"?

4. Wasn't ArtForz mining bitcoins with his GPU miner before any publicly released GPU miner existed? How is that any different (if it is true, which you have no proof of)?

5. "Nothing can be ASIC resistant". Well what you are doing is taking statements that are made by people who dont have a clue how litecoin works and using that to portray it as if every litecoin support believes that.

Do us all a set of favors:

1. Stop being a Douche.

2. Stop spreading bullshit and claiming that all LTC supporters  agree with certain statements.

3. Grow the fuck up. Most of the other early adopters/devs aren't as fucking childish as you are.

My rant has ended. Have a nice day!

 Cheesy Cheesy Cheesy
Fatal mistake. Don't respond to trolls. The one thing that is MOST effective against them is ignoring them, or at most, talking about them in the 3rd person as though they're not present.

Luke-jr has failed to accept one of the universal truths: If you end up having a problem with the whole world, then perhaps it's not the whole world at fault.

Words of wisdom.

Thanks.

███████████████████████████████████████

            ,╓p@@███████@╗╖,           
        ,p████████████████████N,       
      d█████████████████████████b     
    d██████████████████████████████æ   
  ,████²█████████████████████████████, 
 ,█████  ╙████████████████████╨  █████y
 ██████    `████████████████`    ██████
║██████       Ñ███████████`      ███████
███████         ╩██████Ñ         ███████
███████    ▐▄     ²██╩     a▌    ███████
╢██████    ▐▓█▄          ▄█▓▌    ███████
 ██████    ▐▓▓▓▓▌,     ▄█▓▓▓▌    ██████─
           ▐▓▓▓▓▓▓█,,▄▓▓▓▓▓▓▌          
           ▐▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▌          
    ▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓─  
     ²▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓╩    
        ▀▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓▀       
           ²▀▀▓▓▓▓▓▓▓▓▓▓▓▓▀▀`          
                   ²²²                 
███████████████████████████████████████

. ★☆ WWW.LEALANA.COM        My PGP fingerprint is A764D833.        SMOOTHIE'S HEALTH AND FITNESS JOURNAL          History of Monero development Visualization ★☆ .
LEALANA  PHYSICAL MONERO COINS 999 FINE SILVER.
 
-ck
Staff
Legendary
*
Offline Offline

Activity: 2002


Ruu \o/


View Profile WWW
August 01, 2012, 09:10:36 AM
 #587

I seem to be experiencing a problem with the new CGMiner 2.6.1-official that was not affecting 2.5.0.Scrypt_beta_ee

CGMiner now seems to be seeing cards at times as sick and the program crashes. It has happened to each of my miners at least once since I switched from the 'EE' Beta 2.5.0 yesterday, to 2.6.1 Official.

The last beta ('EE') seemed to run fine for multiple days, but I have yet to experience a SOLID 24 hours from 2.6.1-Official.....without a crash.

The 'EE' version, for reference...
win32: http://www.mediafire.com/?0yxa1cf35pk5q2i

Latest Commits included!

You know it's interesting, I'm in the middle of tracking down the cause of this exact issue for someone else. It turns out that the change to avoid flooding pools on longpoll means I limit outgoing connections - this means there is more likely to be a period a GPU is idle while waiting for work. Now what's interesting is that short periods of idle followed by huge bursts of activity again tend to make cards more unstable than running them flat out 100% of the time.  Some testing showed the cards needed to have their clocks decreased quite a lot to be able to tolerate this. So overclocking which was fine before suddenly becomes unstable. I'm experimenting with a feature I call no-idle which gives GPUs fake work to keep them busy to see if this improves stability, but it's still early stages. If you have more stability with the scrypt patched 2.5.0, then by all means go with that for now.

Primary developer/maintainer for cgminer and ckpool/ckproxy.
Pooled mine at kano.is, solo mine at solo.ckpool.org
-ck
bitlane
Internet detective
Sr. Member
****
Offline Offline

Activity: 462


I heart thebaron


View Profile
August 01, 2012, 10:57:25 AM
 #588

I seem to be experiencing a problem with the new CGMiner 2.6.1-official that was not affecting 2.5.0.Scrypt_beta_ee

CGMiner now seems to be seeing cards at times as sick and the program crashes. It has happened to each of my miners at least once since I switched from the 'EE' Beta 2.5.0 yesterday, to 2.6.1 Official.

The last beta ('EE') seemed to run fine for multiple days, but I have yet to experience a SOLID 24 hours from 2.6.1-Official.....without a crash.

The 'EE' version, for reference...
win32: http://www.mediafire.com/?0yxa1cf35pk5q2i

Latest Commits included!

You know it's interesting, I'm in the middle of tracking down the cause of this exact issue for someone else. It turns out that the change to avoid flooding pools on longpoll means I limit outgoing connections - this means there is more likely to be a period a GPU is idle while waiting for work. Now what's interesting is that short periods of idle followed by huge bursts of activity again tend to make cards more unstable than running them flat out 100% of the time.  Some testing showed the cards needed to have their clocks decreased quite a lot to be able to tolerate this. So overclocking which was fine before suddenly becomes unstable. I'm experimenting with a feature I call no-idle which gives GPUs fake work to keep them busy to see if this improves stability, but it's still early stages. If you have more stability with the scrypt patched 2.5.0, then by all means go with that for now.

Could this problem simply be a bi-product of the switch to displaying MH/s while SCRYPT mining in 2.6.1 ?

In the 2.5.0 'EE' SCRYPT Beta, CGMiner displayed KH/s exclusively (GPU, Rig Summary/5s/AVG & API).
In 2.6.1, while SCRYPT mining, CGMiner seems to mix the 2 scales using KH/s for individual GPUs and then MH/s for Rig Summary/5s/AVG & API output.

Isn't this item the most significant difference that has changed from 2.5.0beta to 2.6.1official ? (for SCRYPT mining anyway).

Could the 'sickness' simply be a low value output in MH/s, causing CGMiner to panic ?

Would it be possible that when in --SCRYPT mode that the units displayed are 100% in KH/s ? (GPU, Rig Summary/5s/AVG & API output ?)
Or, perhaps have the option to specify in the CMD line using something like [ --units khash] or [--units mhash] ?

...Just a thought.
thanks again,
bitlane.



BlackBison
Sr. Member
****
Offline Offline

Activity: 253



View Profile
August 01, 2012, 11:01:54 AM
 #589

so has anyone tested cgminer with 5850s? is it faster than reaper and therefore worth me switching over, or is it roughly the the same hashrate?

EskimoBob
Legendary
*
Offline Offline

Activity: 910


Quality Printing Services by Federal Reserve Bank


View Profile
August 01, 2012, 11:28:25 AM
 #590

...
Would it be possible that when in --SCRYPT mode that the units displayed are 100% in KH/s ? (GPU, Rig Summary/5s/AVG & API output ?)
Or, perhaps have the option to specify in the CMD line using something like [ --units khash] or [--units mhash] ?

As you probably have noticed, there is no space for the Kh/s in the upper row when you are in 1000 range Smiley
Mine shows Kh/s  

While reading what I wrote, use the most friendliest and relaxing voice in your head.
BTW, Things in BTC bubble universes are getting ugly....
bitlane
Internet detective
Sr. Member
****
Offline Offline

Activity: 462


I heart thebaron


View Profile
August 01, 2012, 11:47:23 AM
 #591

...
Would it be possible that when in --SCRYPT mode that the units displayed are 100% in KH/s ? (GPU, Rig Summary/5s/AVG & API output ?)
Or, perhaps have the option to specify in the CMD line using something like [ --units khash] or [--units mhash] ?

As you probably have noticed, there is no space for the Kh/s in the upper row when you are in 1000 range Smiley
Mine shows Kh/s  

You sure about THAT ?



<EDIT>

My API Web App displays whatever it's fed and does so automatically (kh rather than the MH as I showed above)...
(although values are dynamic, units are static in the app and need to be changed....not a big deal)


burnside
Hero Member
*****
Offline Offline

Activity: 882



View Profile WWW
August 01, 2012, 03:04:34 PM
 #592

so if i am not mistaken the main difference between LTC and BTC was the fact that LTC was targeted at CPU mining only.

now that you can mine it on GPU, arent both currencies basically the same? making one redundant?

Nah, no FPGA's or ASIC's for LTC. (thus still accessible by the common man) LTC will have more coins in the wild.  Faster confirms.  LTC ftw.

edit... delayed reaction.  doh.

I'm not a Coinbase fan -- I placed a buy order, they took the funds out of my account, then a week later the price went up and they canceled the buy and closed my account.  You've been warned.  Use a different exchange.
-ck
Staff
Legendary
*
Offline Offline

Activity: 2002


Ruu \o/


View Profile WWW
August 02, 2012, 12:08:50 AM
 #593

Could the 'sickness' simply be a low value output in MH/s, causing CGMiner to panic ?
No, but that bug WAS in 2.6.0 which was the main reason for 2.6.1

We have found that this problem you describe happens with driver 12.6 when it doesn't happen with the earlier drivers which were better (specifically the so-called 8.921 release which was the first beta driver supporting 7970s).

Primary developer/maintainer for cgminer and ckpool/ckproxy.
Pooled mine at kano.is, solo mine at solo.ckpool.org
-ck
VelvetLeaf
Member
**
Offline Offline

Activity: 98


View Profile
August 02, 2012, 07:50:23 AM
 #594

That means the settings aren't at best. Play with it, try lower concurrencys and higher ones. The settings won't be the same as reaper. Try other worksizes.

2x5970, got 1022 kH/s with this parameter :
--scrypt --thread-concurrency 6144 --intensity 13 --lookup-gap 2 --gpu-threads 1 --gpu-memclock 1100 --gpu-engine 775 --auto-fan --temp-target 75 --worksize 256

Still less than 1.2 MH/s from reaper with the same configuration.
But temperature so far is good, better than reaper's temperature.

EDIT : Got 1.2 MH/s with intensity 15
I'm afraid to raise intensity because of the warning in readme : "Intensities over 13 start writing over the same ram and it is highly dependent on the GPU, but they can start actually DECREASING your hashrate, or even worse, start producing garbage with rejects skyrocketing."

EDIT 2 : Still fluctuating, now at 1.0 MH/s average.

BTC : 1GN81dxzxyFPQsyAtdocXr5S9Mcg4wcfFG
LTC : LgmYvXsYXc4xdjsMKXJWqtagxVvioK6iaw
FC : 6dpSnKMtttUUYzaRu1EB7Lu18PBRVHU3V7
SAC
Sr. Member
****
Offline Offline

Activity: 322


View Profile
August 02, 2012, 08:28:40 AM
 #595

That means the settings aren't at best. Play with it, try lower concurrencys and higher ones. The settings won't be the same as reaper. Try other worksizes.

2x5970, got 1022 kH/s with this parameter :
--scrypt --thread-concurrency 6144 --intensity 13 --lookup-gap 2 --gpu-threads 1 --gpu-memclock 1100 --gpu-engine 775 --auto-fan --temp-target 75 --worksize 256

Still less than 1.2 MH/s from reaper with the same configuration.
But temperature so far is good, better than reaper's temperature.

EDIT : Got 1.2 MH/s with intensity 15
I'm afraid to raise intensity because of the warning in readme : "Intensities over 13 start writing over the same ram and it is highly dependent on the GPU, but they can start actually DECREASING your hashrate, or even worse, start producing garbage with rejects skyrocketing."

EDIT 2 : Still fluctuating, now at 1.0 MH/s average.

2x5970, gets me 1388 kH/s, 347Kh/s per core with these parameters :
--scrypt --thread-concurrency 8000 --intensity 18 --lookup-gap 2 --gpu-threads 2 --gpu-memclock 1050 --gpu-engine 750 --auto-fan --auto-gpu --temp-target 81 --worksize 256

Stales at ~1-3% depending on how fast the network is running as most of them come on the new blocks.
VelvetLeaf
Member
**
Offline Offline

Activity: 98


View Profile
August 02, 2012, 08:36:15 AM
 #596

2x5970, gets me 1388 kH/s, 347Kh/s per core with these parameters :
--scrypt --thread-concurrency 8000 --intensity 18 --lookup-gap 2 --gpu-threads 2 --gpu-memclock 1050 --gpu-engine 750 --auto-fan --auto-gpu --temp-target 81 --worksize 256

Stales at ~1-3% depending on how fast the network is running as most of them come on the new blocks.

Got cgminer isn't responding. What's your driver version ?

BTC : 1GN81dxzxyFPQsyAtdocXr5S9Mcg4wcfFG
LTC : LgmYvXsYXc4xdjsMKXJWqtagxVvioK6iaw
FC : 6dpSnKMtttUUYzaRu1EB7Lu18PBRVHU3V7
SAC
Sr. Member
****
Offline Offline

Activity: 322


View Profile
August 02, 2012, 04:27:52 PM
 #597

2x5970, gets me 1388 kH/s, 347Kh/s per core with these parameters :
--scrypt --thread-concurrency 8000 --intensity 18 --lookup-gap 2 --gpu-threads 2 --gpu-memclock 1050 --gpu-engine 750 --auto-fan --auto-gpu --temp-target 81 --worksize 256

Stales at ~1-3% depending on how fast the network is running as most of them come on the new blocks.

Got cgminer isn't responding. What's your driver version ?

Ubuntu 11.04 12.6 display driver SDK 2.7.
pekv2
Hero Member
*****
Offline Offline

Activity: 770



View Profile
August 03, 2012, 07:35:27 AM
 #598

Is there a way to implement an option to add thread-concurrency configuration through cgminer.exe?

Everytime I do a [W], cgminer overwrites my config file, removing

Code:
"thread-concurrency" : "7200,7200",

from cgminer.conf.

Or is it already there and I am overlooking it?
kano
Legendary
*
Offline Offline

Activity: 1932


Linux since 1997 RedHat 4


View Profile
August 03, 2012, 07:38:32 AM
 #599

Is there a way to implement an option to add thread-concurrency configuration through cgminer.exe?

Everytime I do a [W], cgminer overwrites my config file, removing

Code:
"thread-concurrency" : "7200,7200",

from cgminer.conf.

Or is it already there and I am overlooking it?
https://github.com/ckolivas/cgminer/pull/279

Pool: https://kano.is BTC: 1KanoiBupPiZfkwqB7rfLXAzPnoTshAVmb
CKPool and CGMiner developer, IRC FreeNode #ckpool and #cgminer kanoi
Help keep Bitcoin secure by mining on pools with Stratum, the best protocol to mine Bitcoins with ASIC hardware
pekv2
Hero Member
*****
Offline Offline

Activity: 770



View Profile
August 03, 2012, 08:35:03 AM
 #600


Oh, so it is a known problem. mk. Thanks.
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 [30] 31 »
  Print  
 
Jump to:  

Sponsored by , a Bitcoin-accepting VPN.
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!