[ANN] cudaMiner & ccMiner CUDA based mining applications [Windows/Linux/MacOSX] - page 558.

antonio8

legendary

Activity: 1400

Merit: 1000

Quote from: OmarGsPools on April 27, 2014, 06:39:46 PM

Let me know if you guys need any pools for some of the odd algorithms. I run a mini 750Ti GPU farm so instead of solo mining I'd be fine with starting up pools.

I like the look of you jackpot coin pool. Have you tried to get the ccminer implemented into it?

OmarGsPools

sr. member

Activity: 392

Merit: 250

Let me know if you guys need any pools for some of the odd algorithms. I run a mini 750Ti GPU farm so instead of solo mining I'd be fine with starting up pools.

GarlukKY

member

Activity: 81

Merit: 10

Thanks, bigjme, for the Fermi 1 windows compile. My hash rate on Groestl is up 25+% on my 750 TI vs the 0.6 version Grin

bigjme

sr. member

Activity: 350

Merit: 250

Quote from: DELTA9 on April 27, 2014, 05:56:13 PM

Still having trouble mining Myriad with my GTX 750. I am so frustrated, did not think it would be this difficult. Downloading latest nVidia drivers to see if it helps. When I run my command to start ccminer the program runs for 1 second then closes, so I can't even read the error messages. Can someone please share their knowledge with me before I punch another hole in my wall?

run the bat file from command prompt and it will pause long enough for you to see the error

Quote from: hacksaw907 on April 27, 2014, 05:31:53 PM

I would like both if possible, but if that is too much trouble, x64 only would be great.

Thank you kindly for your help!

Compute 3.5 with #define MAXWELL_OR_FERMI 0
https://mega.co.nz/#!BstlBaKJ!kIRU2EVLmnt-Uw6mDt7nwau3Utq4iqM1t9bLTPHzIk4

DELTA9

member

Activity: 102

Merit: 10

Still having trouble mining Myriad with my GTX 750. I am so frustrated, did not think it would be this difficult. Downloading latest nVidia drivers to see if it helps. When I run my command to start ccminer the program runs for 1 second then closes, so I can't even read the error messages. Can someone please share their knowledge with me before I punch another hole in my wall?

bigjme

sr. member

Activity: 350

Merit: 250

Quote from: cbuchner1 on April 27, 2014, 05:51:19 PM

Quote from: bigjme on April 27, 2014, 05:29:08 PM

i am messing with it now and it will most likely be a better solution for you to use then nvml as its 32 and 64 bit

"Initially exposed only to OEMs and game developer tools, NVAPI is now available for download to all developers interested in building Windows applications on NVIDIA GPUs."

it's Windows only. Meh.

Christian

hmm what about implement it into the windows build and use nvidia-smi under linux?
you already have the if definition for if its windows or not

you would need to use something like this

Code:

#ifdef _WIN64
   return "Windows 32-bit";
   #elif _WIN32
   return "Windows 64-bit";
   #elif __unix || __unix__
   return "Unix";
   #elif __APPLE__ || __MACH__
   return "Mac OSX";
   #elif __linux__
   return "Linux";
   #else
   return "Other";
#endif

cbuchner1

hero member

Activity: 756

Merit: 502

Quote from: bigjme on April 27, 2014, 05:29:08 PM

i am messing with it now and it will most likely be a better solution for you to use then nvml as its 32 and 64 bit

"Initially exposed only to OEMs and game developer tools, NVAPI is now available for download to all developers interested in building Windows applications on NVIDIA GPUs."

it's Windows only. Meh.

Christian

bigjme

sr. member

Activity: 350

Merit: 250

Quote from: hacksaw907 on April 27, 2014, 05:31:53 PM

I would like both if possible, but if that is too much trouble, x64 only would be great.

Thank you kindly for your help!

i will compile it for you now then

djm34

legendary

Activity: 1400

Merit: 1050

Quote from: cbuchner1 on April 27, 2014, 05:26:53 PM

jackpotcoin is going to adopt the keccak-only approach permanently. Meaning ccminer will continue working past block 6000.

see line 139ff here
https://github.com/rtc29462/cpuminer/blob/master/jackpot.c

the other hashing functions have been commented out.

I would have fixed it differently, but I wasn't asked Wink

Christian

well it isn't like he invented any of the algo, he doesn't have to feel attached to them...

hacksaw907

newbie

Activity: 2

Merit: 0

Quote from: bigjme on April 27, 2014, 05:11:13 PM

Quote from: hacksaw907 on April 27, 2014, 05:04:27 PM

Would it be possible for you (or someone else) to compile it with the "#define MAXWELL_OR_FERMI" flag to 0? I would like to use the myriad-groestl algo, but my 780's are getting the insane hashrates with no confirmations bug, as others are getting on non-Maxwell cards.

I tried doing it myself, but Nvidia's CUDA toolkit doesn't recognize Visual Studio Express as a legitimate VS application, so I'm not exactly sure what else to do.

If you could even point me in the direction where I can find instructions to compile it myself if compiling it is too much trouble on your end, I'd really appreciate it.

https://mega.co.nz/#!Aw13XIYD!ETSB1GRSQdmqGt1jnlMLV4ZOhJ__qtfOSc0vOMKyUQY
that one does have the flag set to 0, but doesn't have the spelling error fixed, which may be causing errors

so you need it with flag 0, but what version? (x86 or x64)

on a side note, anyone using nvapi, if you can help me figure out how on earth you can return an NvAPI_ShortString value that would be great, i can not for the life of me get it to return the fullname of the card, compiling is throwing and issue, here is the bit of code, and the error

Code:

NvAPI_ShortString name;
syntax error : missing ';' before identifier 'name'

ooo and ccminer does not mine scrypt

I would like both if possible, but if that is too much trouble, x64 only would be great.

Thank you kindly for your help!

bigjme

sr. member

Activity: 350

Merit: 250

Quote from: cbuchner1 on April 27, 2014, 05:26:53 PM

I would have fixed it differently, but I wasn't asked Wink

Christian

you don't have time, your too busy finding errors in coins Wink

christian have you ever looked through the docs for nvapi?

i am messing with it now and it will most likely be a better solution for you to use then nvml as its 32 and 64 bit

cbuchner1

hero member

Activity: 756

Merit: 502

jackpotcoin is going to adopt the keccak-only approach permanently. Meaning ccminer will continue working past block 6000.

see line 139ff here
https://github.com/rtc29462/cpuminer/blob/master/jackpot.c

the other hashing functions have been commented out.

I would have fixed it differently, but I wasn't asked Wink

Christian

bigjme

sr. member

Activity: 350

Merit: 250

Quote from: BitSmacker on April 27, 2014, 05:18:25 PM

Well then this project is completely different than what cudaminer started out as and a new thread should be started unless you are trying to keep this under wraps. Also, does that mean the original cudaminer project has been abandoned?

cudaminer is still working, and this thread is still for cudaminer, we just often talk about ccminer

RavenX

member

Activity: 112

Merit: 10

Quote from: BitSmacker on April 27, 2014, 05:09:43 PM

Is there any improvement for scrypt coins with ccminer vs cudaminer? If not and this is now all about the asic resistant coins then perhaps a new thread should be started.

http://www.cudaminers.net/forum/index.php

Give it a try.
Approved and supported by C&C ~~music~~miner factory... Grin

BitSmacker

member

Activity: 84

Merit: 10

Quote from: bigjme on April 27, 2014, 05:11:13 PM

ooo and ccminer does not mine scrypt

Well then this project is completely different than what cudaminer started out as and a new thread should be started unless you are trying to keep this under wraps. Also, does that mean the original cudaminer project has been abandoned?

bigjme

sr. member

Activity: 350

Merit: 250

Quote from: hacksaw907 on April 27, 2014, 05:04:27 PM

Would it be possible for you (or someone else) to compile it with the "#define MAXWELL_OR_FERMI" flag to 0? I would like to use the myriad-groestl algo, but my 780's are getting the insane hashrates with no confirmations bug, as others are getting on non-Maxwell cards.

I tried doing it myself, but Nvidia's CUDA toolkit doesn't recognize Visual Studio Express as a legitimate VS application, so I'm not exactly sure what else to do.

If you could even point me in the direction where I can find instructions to compile it myself if compiling it is too much trouble on your end, I'd really appreciate it.

https://mega.co.nz/#!Aw13XIYD!ETSB1GRSQdmqGt1jnlMLV4ZOhJ__qtfOSc0vOMKyUQY
that one does have the flag set to 0, but doesn't have the spelling error fixed, which may be causing errors

so you need it with flag 0, but what version? (x86 or x64)

on a side note, anyone using nvapi, if you can help me figure out how on earth you can return an NvAPI_ShortString value that would be great, i can not for the life of me get it to return the fullname of the card, compiling is throwing and issue, here is the bit of code, and the error

Code:

NvAPI_ShortString name;
syntax error : missing ';' before identifier 'name'

ooo and ccminer does not mine scrypt

BitSmacker

member

Activity: 84

Merit: 10

Is there any improvement for scrypt coins with ccminer vs cudaminer? If not and this is now all about the asic resistant coins then perhaps a new thread should be started.

hacksaw907

newbie

Activity: 2

Merit: 0

Quote from: bigjme on April 27, 2014, 08:11:14 AM

Quote from: Dotcommie on April 27, 2014, 06:21:49 AM

Quote from: cbuchner1 on April 27, 2014, 04:12:53 AM

Quote from: cvax on April 27, 2014, 02:16:48 AM

Quote from: bigjme on April 26, 2014, 08:48:48 PM

yeh i can see a lot of people are going to ask, here is 32bit and 64bit ccminer 0.6 compiled
https://mega.co.nz/#!Aw13XIYD!ETSB1GRSQdmqGt1jnlMLV4ZOhJ__qtfOSc0vOMKyUQY

Thanks bigjme. Is this Compute 3.5? I tried using your compile with my same old groestl bat file. The performance was worse. From 3100+ kh/s down to ~2450kh/s. You experiencing any difference?

on Windows this should now be built in four flavours, when done correctly Wink

Compute 2.0 with #define MAXWELL_OR_FERMI 1 <-- Fermi
Compute 3.0 with #define MAXWELL_OR_FERMI 0 <-- old Kepler
Compute 3.5 with #define MAXWELL_OR_FERMI 0 <-- highend Kepler and GT630/640 with compute 3.5
Compute 3.5 with #define MAXWELL_OR_FERMI 1 <-- for Maxwell

the flag must be set in both .cu modules with "groestl" in its name. It affects
the speed optimization of the Groestl code.

at some point I will code a runtime selection of the MAXWELL_OR_FERMI flag.

Christian

I'm guessing the build bigjme made was for maxwell and that's why my 670 in the system is reporting crazy high hashrate, but is there a chance it's actually hashing between 1-2Mhash? lol (I normally run separate bat files but I just left it since it looks like I've missed the train on this one anyways)

I just compiled the build straight from github sorry guys.

Would it be possible for you (or someone else) to compile it with the "#define MAXWELL_OR_FERMI" flag to 0? I would like to use the myriad-groestl algo, but my 780's are getting the insane hashrates with no confirmations bug, as others are getting on non-Maxwell cards.

I tried doing it myself, but Nvidia's CUDA toolkit doesn't recognize Visual Studio Express as a legitimate VS application, so I'm not exactly sure what else to do.

If you could even point me in the direction where I can find instructions to compile it myself if compiling it is too much trouble on your end, I'd really appreciate it.

bigjme

sr. member

Activity: 350

Merit: 250

Quote from: TeaWolf913 on April 27, 2014, 04:37:41 PM

do u have a link to this GPU-Z you speak of, should I just google it

yes just google it

TeaWolf913

newbie

Activity: 5

Merit: 0

Quote from: atc1 on April 27, 2014, 01:43:59 AM

Quote from: TeaWolf913 on April 26, 2014, 02:27:15 PM

Well... After the whole, made this guy, stuff, it says Unable to query CUDA driver version! Is an nVidia Driver installed?

Make sure you have driver version 190.38 installed,because apparently in old cards, the newest drivers are more memory aggressive, reporting less free memory to the system. You'll have to revert back to your older drivers if you're going to use a video card with the bare minimum amount of RAM.

But before that,just download GPU-Z and make sure that the CUDA checkbox at the bottom is ticked. Yes,people have large configs for cards, but that is needed only when you want specific configs,but the autotune is generally a really good option.

Thanks for all the help. I'll try it. But first of all, do u have a link to this GPU-Z you speak of, should I just google it

Topic: [ANN] cudaMiner & ccMiner CUDA based mining applications [Windows/Linux/MacOSX] - page 558. (Read 3426922 times)