[ANN] ccminer 2.3 - opensource - GPL (tpruvot)

[Epsylon3]

not here try --cpu-priority 3

[1714142332]

1714142332

[1714142332]

1714142332

[1714142332]

1714142332

[1714142332]

1714142332

[GoingAround]

Folks, last version is better or not? My version 1.8.3

[Nokia555]

Added support for EthMonitoring, CCMiner 2.0-2.2, Alexis and more forks.

https://ethmonitoring.com

[AlteregoX]

How to change the intensity setting? tried -i and -intensity

[tpd09]

How to change the intensity setting? tried -i and -intensity

its "-i auto" for example

[scryptr]

How to change the intensity setting? tried -i and -intensity

its "-i auto" for example

PLACE A NUMBER AFTER THE "-i"--

For example, "-i 21".  However, CCminer will crash if intensity is set too high.  Leaving it at default (no "-i" flag), or setting it to "-i auto" may be more safe.       

Different CCminer variants will take different settings for intensity.  For example, Nanashi's CCminer code works well with higher values of intensity, but other versions would crash.  Some coders set the default intensity so high that their code crashes with any extra value for intensity.

It takes experimentation to find the best value.       --scryptr

[rodman80]

Epsylon,

Did you try Cuda 9 RC yet? Looks like __shfl was deprecated in favor of something else. There´s no programming guide for V9 yet so I don't know for sure.

I´m trying to give it a shot to see if they improved compiling for sm_61.

Almost all algos run slower on sm_61 compiled code, comparing to sm_52, which is a bit odd.

[ghostwalker.ph]

is it possible to dual mine on ccminer?

for example:

Problem no.1

Rig: 1060 & 1070

1060 mining on timetravel (bitcore)

1070 mining on tribus (denarius)

what is the code for problem no.1?

Problem no. 2

Rig: 1080ti

1080ti mining on timetravel and denarius at the same time.

what is the code for problem no. 2? is this possible?

thank you very much.

[Epsylon3]

Epsylon,

Did you try Cuda 9 RC yet? Looks like __shfl was deprecated in favor of something else. There´s no programming guide for V9 yet so I don't know for sure.

I´m trying to give it a shot to see if they improved compiling for sm_61.

Almost all algos run slower on sm_61 compiled code, comparing to sm_52, which is a bit odd.

yes... sm_61 is an emulation of sm_60 P100 fp16 stuff, and when used (look like it is sometimes) its slower than normal ops generated for sm_5x ... So nothing changed for the GTX (and mining cards) about that since CUDA 8

ghostwalker: -d 0 or -d 1

[Lumikor]

Epsylon,

Did you try Cuda 9 RC yet? Looks like __shfl was deprecated in favor of something else. There´s no programming guide for V9 yet so I don't know for sure.

I´m trying to give it a shot to see if they improved compiling for sm_61.

Almost all algos run slower on sm_61 compiled code, comparing to sm_52, which is a bit odd.

yes... sm_61 is an emulation of sm_60 P100 fp16 stuff, and when used (look like it is sometimes) its slower than normal ops generated for sm_5x ... So nothing changed for the GTX (and mining cards) about that since CUDA 8

ghostwalker: -d 0 or -d 1

So when compiling for 1080ti, you recommend using sm_60 or 61?

[tpd09]

How to change the intensity setting? tried -i and -intensity

its "-i auto" for example

PLACE A NUMBER AFTER THE "-i"--

For example, "-i 21".  However, CCminer will crash if intensity is set too high.  Leaving it at default (no "-i" flag), or setting it to "-i auto" may be more safe.       

Different CCminer variants will take different settings for intensity.  For example, Nanashi's CCminer code works well with higher values of intensity, but other versions would crash.  Some coders set the default intensity so high that their code crashes with any extra value for intensity.

It takes experimentation to find the best value.       --scryptr

@scryptr - when Im using high intensity values like "25", the power limit is very unstable, but it does generate higher hashes. what is the implications of such power variances?
thanks

[scryptr]

How to change the intensity setting? tried -i and -intensity

its "-i auto" for example

PLACE A NUMBER AFTER THE "-i"--

For example, "-i 21".  However, CCminer will crash if intensity is set too high.  Leaving it at default (no "-i" flag), or setting it to "-i auto" may be more safe.       

Different CCminer variants will take different settings for intensity.  For example, Nanashi's CCminer code works well with higher values of intensity, but other versions would crash.  Some coders set the default intensity so high that their code crashes with any extra value for intensity.

It takes experimentation to find the best value.       --scryptr

@scryptr - when Im using high intensity values like "25", the power limit is very unstable, but it does generate higher hashes. what is the implications of such power variances?
thanks

LOWER YOUR INTENSITY--

Stability and efficiency are just as important as maximum hash rate.  A higher, unstable hash rate is wasting wattage and time.       --scryptr

[rodman80]

yes... sm_61 is an emulation of sm_60 P100 fp16 stuff, and when used (look like it is sometimes) its slower than normal ops generated for sm_5x ... So nothing changed for the GTX (and mining cards) about that since CUDA 8

ghostwalker: -d 0 or -d 1

Oh, I didn't know that. It does make sense the poor performance then, since fp16 in P100 is a lot different than in GTX family. I hope they change the emulation somehow in the future.

I'm trying to learn the new tricks using Cuda9 RC. I think the cooperative groups approach might be interesting for a few algos.

I'll send you a PM if I discover anything useful for ccminer, but don't expect too much, since I don't have even 20% of your coding expertise

Congrats and keep up the good work.

Rod

[Christofu450]

Hi guys,

I'm trying to get ccminer to work on Ubuntu 16.04 with a GTX 1080.
I've installed successfully latest Nvidia driver from here: http://www.nvidia.fr/download/driverResults.aspx/121202.
I've installed successfully cuda toolkit 8.0 from here: https://developer.nvidia.com/cuda-downloads.

I have compiled and installed succecssfully ccminer using sm_61 (see: https://github.com/tpruvot/ccminer/wiki/Compatibility).

I've tried many things but I always end up with this when trying to run ccminer:

Code:

*** ccminer 2.2 for nVidia GPUs by tpruvot@github ***
    Built with the nVidia CUDA Toolkit 8.0 64-bits

  Originally based on Christian Buchner and Christian H. project
  Include some algos from alexis78, djm34, sp, tsiv and klausT.

BTC donation address: 1AJdfCpLWPNoAMDfHF1wD5y8VgKSSTHxPo (tpruvot)

[2017-08-04 21:29:08] Your system does not support CUDA 8.0 API!

But my system should support cuda 8.0 of course.

Does anyone have a clue ?
Please tell me if you need more details.

EDIT: found the issue, I added /usr/local/cuda/lib64/stubs to ld path which made a library loaded twice, and once from the wrong path.
I removed that path the same way I added it:

Code:

sudo rm /etc/ld.so.conf.d/cuda64-stubs.conf

Et voilà!

[yunihuni]

I guess CUDA 9 released.

https://developer.nvidia.com/cuda-toolkit/whatsnew

[pdiogo]

Hello all,

I've been trying to mine SIGT using the latest skunk algo, but I can't seem to do it on Ubuntu 16.06 with a Nvidia Tesla K80:

Code:

pdiogo@nc6:~/temp/ccminer$ ccminer -a skunk -o stratum+tcp://sigt.pool.mn:8732 -u hidden -p hidden
*** ccminer 2.2 for nVidia GPUs by tpruvot@github ***
    Built with the nVidia CUDA Toolkit 8.0 64-bits

  Originally based on Christian Buchner and Christian H. project
  Include some algos from alexis78, djm34, sp, tsiv and klausT.

BTC donation address: 1AJdfCpLWPNoAMDfHF1wD5y8VgKSSTHxPo (tpruvot)

[2017-08-04 21:53:37] Starting on stratum+tcp://sigt.pool.mn:8732
[2017-08-04 21:53:37] NVML GPU monitoring enabled.
[2017-08-04 21:53:37] 1 miner thread started, using 'skunk' algorithm.
[2017-08-04 21:53:38] Stratum difficulty set to 0.00390625
[2017-08-04 21:53:38] GPU #0: Intensity set to 17, 131072 cuda threads
Cuda error in func 'skein512_cpu_setBlock_80' at line 1095 : invalid device symbol.

Any help?

EDIT: Nvidia drivers and CUDA is ok, as I'm able to mine with different miners. I also did add the correct nvcc_ARCH flag, according to https://github.com/tpruvot/ccminer/wiki/Compatibility

[the.vlado]

Finally I manage to compile ccminer. I needed two days to install correct version of VS  (2013) and to be able to compile it without errors.

But when I run it from VS, it will only connect to the pool and receive blocks. Card is at 0% and mining is not happening. What should I do?

Thank you.

[Larz952002]

anyone else getting a ton of "illegal memory" errors?  is there a fix?

[lexele]

anyone else getting a ton of "illegal memory" errors?  is there a fix?

It's usually because of too much overclocking

[Strebba90]

i try to compile on Ubuntu 16.04 but display this error at the and of make.

ccminer-ccminer.o: nella funzione "miner_thread(void*)":
ccminer.cpp:(.text+0x7b75): riferimento non definito a "scanhash_timetravel10"
ccminer-util.o: nella funzione "print_hash_tests":
util.cpp:(.text+0x7a29): riferimento non definito a "timetravel10_hash"
ccminer-bench.o: nella funzione "algo_free_all":
bench.cpp:(.text+0x215): riferimento non definito a "free_timetravel10"
collect2: error: ld returned 1 exit status
Makefile:1128: set di istruzioni per l'obiettivo "ccminer" non riuscito
make[2]: *** [ccminer] Errore 1
make[2]: uscita dalla directory "/home/rig/miners/ccminer-palginpav"
Makefile:2113: set di istruzioni per l'obiettivo "all-recursive" non riuscito
make[1]: *** [all-recursive] Errore 1
make[1]: uscita dalla directory "/home/rig/miners/ccminer-palginpav"
Makefile:631: set di istruzioni per l'obiettivo "all" non riuscito
make: *** [all] Errore 2

anyone can help me?