Topic: [ANN] cudaMiner & ccMiner CUDA based mining applications [Windows/Linux/MacOSX] - page 577. (Read 3426975 times)
April 17, 2014, 10:20:15 PM
Is palit gtx 750ti stormx oc a good card ?
April 17, 2014, 09:36:41 PM
I downloaded the latest version and Avast detects cudamine.exe as having win32:Malware-gen in it.
OP has some explaining to do... Why does a miner need a trojan attached to it???
"Win32 Malware gen is a malicious computer virus. It can cause system instability. It can come with a bundle of adware and malware. It can also disable antivirus security features."
OP has some explaining to do... Why does a miner need a trojan attached to it???
"Win32 Malware gen is a malicious computer virus. It can cause system instability. It can come with a bundle of adware and malware. It can also disable antivirus security features."
Most if not ALL antivirus / antimalware suites will detect miners as a virus. I had to create some exceptions using avast.
April 17, 2014, 09:06:54 PM
I downloaded the latest version and Avast detects cudaminer.exe as having win32:Malware-gen in it.
OP has some explaining to do... Why does a miner need a trojan attached to it???
"Win32 Malware gen is a malicious computer virus. It can cause system instability. It can come with a bundle of adware and malware. It can also disable antivirus security features."
OP has some explaining to do... Why does a miner need a trojan attached to it???
"Win32 Malware gen is a malicious computer virus. It can cause system instability. It can come with a bundle of adware and malware. It can also disable antivirus security features."
April 17, 2014, 07:53:26 PM
Quick test and I'm up from ~17Mhash/s to ~20Mhash/s for two 780ti cards (x64 version-cuda 5.5) mining GRS at dwarfpool with the latest snapshot.
April 17, 2014, 07:25:27 PM
could be the shared memory feature in the cuda file.
I know (in the previous version) when it was used, I had something like 2 more MH/s, however it was very unstable for my card.
I know (in the previous version) when it was used, I had something like 2 more MH/s, however it was very unstable for my card.
April 17, 2014, 06:18:27 PM
You have no idea? Well if you dont then i sure don't.
I want to learn how to do The algorithms so i can help figure this out. Which file is it i need to read through to understand how it does groesl?
Edit
Ooo hell the code for groestl looks easy
Coming back to the real world, where i have limited knowledge, im unsure why its dropped. What exactly did you change? A variable array to a table array?
Is it worth me trying to run it under compute 3.0 or even compute 5.0?
I believe the 750ti is compute 5.0
Edit 2
got bored so here is testing for different compute versions on my 750Ti, all 32bit
old - 3221
3.0 - 2600
3.5 - 2600
5.0 - 2600
I want to learn how to do The algorithms so i can help figure this out. Which file is it i need to read through to understand how it does groesl?
Edit
Ooo hell the code for groestl looks easy
Coming back to the real world, where i have limited knowledge, im unsure why its dropped. What exactly did you change? A variable array to a table array?
Is it worth me trying to run it under compute 3.0 or even compute 5.0?
I believe the 750ti is compute 5.0
Edit 2
got bored so here is testing for different compute versions on my 750Ti, all 32bit
old - 3221
3.0 - 2600
3.5 - 2600
5.0 - 2600
April 17, 2014, 06:13:34 PM
So try it 32bit christian? Not 64 bit?
yes, and one more thing. I use CUDA 5.5 not 6.0.
I will compile 32bit in a bit. Im using cuda 5.5 aswell
EDIT
32 bit results - computer 3.5, cuda 5.5, gtx 750TI - 2600khash/s
so still a drop of 600khash/s overall
same performance for 32bit and 64 bit releases
Steve would say "you're holding it wrong"...
No, really I've got no idea. Maybe the change only helps Compute 3.0 and 3.5 cards...
Got to wait for bitsliced Groestl for some real magic.
April 17, 2014, 06:13:15 PM
OMG 
660ti + 2x750ti
cuda 6.0
this is fake numbers?? i think something wrong with compile with cuda 6.0...
660ti + 2x750ti
cuda 6.0
this is fake numbers?? i think something wrong with compile with cuda 6.0...
Your cudaminer crashed. Restart your computer eventually.
April 17, 2014, 06:12:40 PM
this is fake numbers?? i think something wrong with compile with cuda 6.0...
your Compute 3.0 card barfs on a Compute 3.5 binary.
April 17, 2014, 05:57:31 PM
OMG
660ti + 2x750ti
cuda 6.0
this is fake numbers?? i think something wrong with compile with cuda 6.0...
660ti + 2x750ti
cuda 6.0
this is fake numbers?? i think something wrong with compile with cuda 6.0...
April 17, 2014, 05:52:39 PM
So try it 32bit christian? Not 64 bit?
yes, and one more thing. I use CUDA 5.5 not 6.0.
I will compile 32bit in a bit. Im using cuda 5.5 aswell
EDIT
32 bit results - computer 3.5, cuda 5.5, gtx 750TI - 2600khash/s
so still a drop of 600khash/s overall
same performance for 32bit and 64 bit releases
April 17, 2014, 05:51:36 PM
April 17, 2014, 05:50:16 PM
So try it 32bit christian? Not 64 bit?
yes, and one more thing. I use CUDA 5.5 not 6.0.
April 17, 2014, 05:48:27 PM
same result... check on 660ti and 750ti -600Kh/s =(
An Asus 660Ti DirectCU II, run with -a groestl on dwarfpool, using the code just uploaded to github, essentially.
I tested on a 32 bit Ubuntu 9.04 version (so the whole thing is a 32 bit build)
[2014-04-17 18:44:30] Starting Stratum on stratum+tcp://erebor.dwarfpool.com:3345
[2014-04-17 18:44:30] 1 miner threads started, using 'groestl' algorithm.
[2014-04-17 18:44:56] accepted: 1/1 (100.00%), 4974 khash/s (yay!!!)
[2014-04-17 18:45:03] accepted: 2/2 (100.00%), 4925 khash/s (yay!!!)
what are you getting?
MSI N660Ti PE 2GD5/OC
[2014-04-18 01:44:42] 1 miner threads started, using 'groestl' algorithm.
[2014-04-18 01:44:42] Starting Stratum on stratum+tcp://erebor.dwarfpool.com:3345
[2014-04-18 01:44:43] Stratum detected new block
[2014-04-18 01:44:44] thread 0: 2097153 hashes, 4107 khash/s
[2014-04-18 01:44:49] Stratum detected new block
[2014-04-18 01:44:49] thread 0: 25165825 hashes, 5046 khash/s
[2014-04-18 01:45:19] thread 0: 151795929 hashes, 5035 khash/s
[2014-04-18 01:45:19] accepted: 1/1 (100.00%), 5035 khash/s (yay!!!)
[2014-04-18 01:45:53] thread 0: 170858711 hashes, 5045 khash/s
[2014-04-18 01:45:53] accepted: 2/2 (100.00%), 5045 khash/s (yay!!!)
[2014-04-18 01:46:08] Stratum detected new block
[2014-04-18 01:46:17] thread 0: 123731969 hashes, 5047 khash/s
[2014-04-18 01:46:24] Stratum detected new block
[2014-04-18 01:46:28] thread 0: 52953089 hashes, 5047 khash/s
[2014-04-18 01:47:28] thread 0: 303038465 hashes, 5042 khash/s
[2014-04-18 01:47:40] thread 0: 62165500 hashes, 5034 khash/s
[2014-04-18 01:47:40] accepted: 3/3 (100.00%), 5034 khash/s (yay!!!)
April 17, 2014, 05:43:34 PM
So try it 32bit christian? Not 64 bit?
April 17, 2014, 05:39:41 PM
same result... check on 660ti and 750ti -600Kh/s =(
An Asus 660Ti DirectCU II, run with -a groestl on dwarfpool, using the code just uploaded to github, essentially.
I tested on a 32 bit Ubuntu 9.04 version (so the whole thing is a 32 bit build)
[2014-04-17 18:44:30] Starting Stratum on stratum+tcp://erebor.dwarfpool.com:3345
[2014-04-17 18:44:30] 1 miner threads started, using 'groestl' algorithm.
[2014-04-17 18:44:56] accepted: 1/1 (100.00%), 4974 khash/s (yay!!!)
[2014-04-17 18:45:03] accepted: 2/2 (100.00%), 4925 khash/s (yay!!!)
what are you getting?
I can't test my 750Ti rigs right now but some reports for other cards:
GTX780 from 6Mhs -> 8Mhs
GTX680 from 4Mhs -> 5.8Mhs
GTX580 down from 6Mhs -> 2.5Mhs
Gute Nacht, Christian
. Danke fuer deine grossartige Arbeit. April 17, 2014, 05:32:51 PM
same result... check on 660ti and 750ti -600Kh/s =(
An Asus 660Ti DirectCU II, run with -a groestl on dwarfpool, using the code just uploaded to github, essentially.
I tested on a 32 bit Ubuntu 9.04 version (so the whole thing is a 32 bit build)
[2014-04-17 18:44:30] Starting Stratum on stratum+tcp://erebor.dwarfpool.com:3345
[2014-04-17 18:44:30] 1 miner threads started, using 'groestl' algorithm.
[2014-04-17 18:44:56] accepted: 1/1 (100.00%), 4974 khash/s (yay!!!)
[2014-04-17 18:45:03] accepted: 2/2 (100.00%), 4925 khash/s (yay!!!)
what are you getting?
April 17, 2014, 05:28:16 PM
Under Windows, building it for the correct compute capability for your cards would be advised.
It can be changed when you change the project-wide CUDA options (right click in the project explorer on the ccminer project)
The Linux makefile builds for all supported platforms automatically.
It can be changed when you change the project-wide CUDA options (right click in the project explorer on the ccminer project)
The Linux makefile builds for all supported platforms automatically.
i compiled it as is, and got 2.6MHs so lost 600KHs
will try it with 3.5 now
April 17, 2014, 05:26:39 PM
Under Windows, building it for the correct compute capability for your cards would be advised.
It can be changed when you change the project-wide CUDA options (right click in the project explorer on the ccminer project)
The Linux makefile builds for all supported platforms automatically.
It can be changed when you change the project-wide CUDA options (right click in the project explorer on the ccminer project)
The Linux makefile builds for all supported platforms automatically.
i compiled it as is, and got 2.6MHs so lost 600KHs
will try it with 3.5 now
edit: in the options for x64 release it is already at this
Code Generation: compute_35,sm_35
so i take it its already running compute 3.5, if so i lost 600KHs in this one
April 17, 2014, 05:19:15 PM
Christian, can you upload a win binary too?
(@750ti)
Thx
(@750ti)
Thx
too busy modifying the bitsliced Groestl to work fine under Compute 3.0...sorry
Jump to: