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

[Epsylon3]

just try --no-gbt

i didnt test but its optional/recommended to check the current block height before the submit of found nonces (to prevent some booo)

[IMweasel]

That works! It removes the error and i think there is work being done. But in normal mode,you still don't see anything happening.
Using -D again,i can see stuff like this going on;

Code:

[2015-05-29 19:24:20] GPU #0: n=58
[2015-05-29 19:24:20] GPU #1: n=80000057
[2015-05-29 19:25:09] GPU #1: n=80001657
[2015-05-29 19:25:09] GPU #0: n=1658
[2015-05-29 19:25:58] GPU #1: n=80002c57
[2015-05-29 19:25:58] GPU #0: n=2c58

When i terminate the batch job,ccminer crashes. But i can still answer the question;

Code:

Terminate batch job (Y/N)?

When i choose Y, i could get a storm of messages like this;

Code:

[2015-05-29 19:29:34] CTRL_BREAK_EVENT received, exiting
[2015-05-29 19:29:36] GPU #0: start=00000319 end=00100318 range=000fffff
[2015-05-29 19:29:36] GPU #0: n=371
[2015-05-29 19:29:36] GPU #0: GeForce GTX 660, 88.00 kH/s
[2015-05-29 19:29:36] GPU #0: start=00000372 end=00100371 range=000fffff
[2015-05-29 19:29:36] GPU #0: n=3ca
[2015-05-29 19:29:36] GPU #0: GeForce GTX 660, 44.00 kH/s
[2015-05-29 19:29:36] GPU #0: start=000003cb end=001003ca range=000fffff
[2015-05-29 19:29:36] GPU #0: n=423
[2015-05-29 19:29:36] GPU #0: GeForce GTX 660, 88.00 kH/s
[2015-05-29 19:29:36] GPU #0: start=00000424 end=00100423 range=000fffff
[2015-05-29 19:29:36] GPU #0: n=47c
[2015-05-29 19:29:36] GPU #0: GeForce GTX 660, 88.00 kH/s
[2015-05-29 19:29:36] GPU #0: start=0000047d end=0010047c range=000fffff
[2015-05-29 19:29:36] GPU #0: n=4d5
[2015-05-29 19:29:36] GPU #0: GeForce GTX 660, 88.00 kH/s
[2015-05-29 19:29:36] GPU #0: start=000004d6 end=001004d5 range=000fffff
[2015-05-29 19:29:36] GPU #0: n=52e
[2015-05-29 19:29:36] GPU #0: GeForce GTX 660, 88.00 kH/s
[2015-05-29 19:29:36] GPU #0: start=0000052f end=0010052e range=000fffff

I noticed that GPU 1 is not mentioned,don't know if this is important!
Would greatly appreciate a bit more help!

edit:both gpu's are 100% utilised,so the working part is Ok,just not the status/output part!

[Epsylon3]

jane is very "slow" at this factor you have to be really patient... generally faster after a few minutes (only tested with -a scrypt-jane:CACH)

[IMweasel]

So... if it keeps doing this, for an hour now, is ccminer actually mining or not?
Because i get Cudaminer to go in a matter of seconds,and it could have had 1, or even 2 blocks if i were lucky by now!

Code:

[2015-05-29 19:39:54] GPU #0: n=58
[2015-05-29 19:40:43] GPU #1: n=80001657
[2015-05-29 19:40:43] GPU #0: n=1658
[2015-05-29 19:41:32] GPU #1: n=80002c57
[2015-05-29 19:41:32] GPU #0: n=2c58
[2015-05-29 19:42:21] GPU #1: n=80004257
[2015-05-29 19:42:21] GPU #0: n=4258
[2015-05-29 19:43:10] GPU #1: n=80005857
[2015-05-29 19:43:10] GPU #0: n=5858
[2015-05-29 19:43:59] GPU #1: n=80006e57
[2015-05-29 19:43:59] GPU #0: n=6e58
[2015-05-29 19:44:48] GPU #1: n=80008457
[2015-05-29 19:44:48] GPU #0: n=8458
[2015-05-29 19:45:37] GPU #1: n=80009a57
[2015-05-29 19:45:37] GPU #0: n=9a58
[2015-05-29 19:46:26] GPU #1: n=8000b057

*CUT*

[2015-05-29 20:33:50] GPU #1: n=8005ac57
[2015-05-29 20:33:54] GPU #0: n=5ac58
[2015-05-29 20:34:39] GPU #1: n=8005c257
[2015-05-29 20:34:43] GPU #0: n=5c258
[2015-05-29 20:35:28] GPU #1: n=8005d857
[2015-05-29 20:35:32] GPU #0: n=5d858
[2015-05-29 20:36:17] GPU #1: n=8005ee57
[2015-05-29 20:36:21] GPU #0: n=5ee58
[2015-05-29 20:37:06] GPU #1: n=80060457
[2015-05-29 20:37:10] GPU #0: n=60458
[2015-05-29 20:37:55] GPU #1: n=80061a57
[2015-05-29 20:38:00] GPU #0: n=61a58
[2015-05-29 20:38:44] GPU #1: n=80063057
[2015-05-29 20:38:49] GPU #0: n=63058
[2015-05-29 20:39:33] GPU #1: n=80064657
[2015-05-29 20:39:38] GPU #0: n=64658

[Epsylon3]

its the purpose of these n= ... its the current scanned nonce

[IMweasel]

Ok,thanks,i'm quite noob at scrypt-jane.
Well well look at what just happened :-)

Code:

[2015-05-29 20:52:11] found => 800799bf bf990780
[2015-05-29 20:52:11] accepted: 0/1 (0.00%), 114.88 H/s booooo

Its a booo though... hope it will straighten out!

Is it normal behaviour that one cannot see the hashing rate when it's working?
Only see it with the block found...

edit: By looking at the hashrate, i realised that CCminer does this job about 20% faster compared to Cudaminer!
(Cudaminer starts out at 110 or 120 but quickly drops and holds at about 90,sometimes 80. CCminer seems to stick at 110)

[Epsylon3]

You will get a lot of boo like that without the gbt check... That means somebody else found it before you. You can sometimes have more details in the wallet debug.log

[IMweasel]

Code:

[2015-05-29 22:42:39] accepted: 0/1 (0.00%), 114.91 H/s booooo
[2015-05-29 22:58:02] GPU #0: GeForce GTX 660, 114.53 H/s
[2015-05-29 22:58:02] accepted: 0/2 (0.00%), 229.44 H/s booooo

[IMweasel]

I noticed that, when i let Cudaminer do it's autotuning, it finds  a way better setting for my cards.
Where CCminer always ends up at k11x1, Cuda starts to work at k68x1 or k34x2. Resulting in being about 155% faster.
When i pass those settings to CCminer,it is perfectly capable to reach the same speeds.
Why is CCminer not capable of finding this setting on its own?

[Epsylon3]

that look like your card memory was used... ive seen that with cudaminer... sometimes you need to restart it more than once to get all the memory available.

i suggest using autotune only after a clean boot... and only then use the found launch config.

Code is almost the same, i made a few changes for the 750ti and 9xx which were handled like a titan in cudaminer, but you are using a GTX 660, so its not that...

[IMweasel]

Using CCminer hashing Neoscypt on GTX660/760 family,mining does not start with error; “sorry neoscrypt is not supported on devices SM 3.0″
Although the changelogs state that it should... shouldn't it?

http://cryptomining-blog.com/?s=ccminer+

1.64-supports Compute 3.0, 3.5, 5.0 and 5.2 GPUs, but note that not all algorithms may work on older GPUs
1.63-supports Compute 3.0, 3.5, 5.0 and 5.2 GPUs

[Epsylon3]

no, read my readme

[bathrobehero]

that look like your card memory was used... ive seen that with cudaminer... sometimes you need to restart it more than once to get all the memory available.

i suggest using autotune only after a clean boot... and only then use the found launch config.

Code is almost the same, i made a few changes for the 750ti and 9xx which were handled like a titan in cudaminer, but you are using a GTX 660, so its not that...

Could be higher device_backoff in ccminer or that device_singlememory is not optional in ccminer?

[liomojo]

I have 5 gt750 ti , they were sleeping for 8 months , cause i was moving to another house, but now are ready for work. Anything that is worth to mine or it is a lost cause now. 

[Epsylon3]

sorry, i cant help for that... ask in the noisy threads

[IMweasel]

no, read my readme

v1.6.3
                  Import and adapt Neoscrypt from djm34 work (SM 5+ only)

Alright... clear.
But, :-)
I've tried ccminer-1.0-edition-djm neoscrypt, and it works just fine... as far as the (noob) eye can tell.
Using the supplied neoscrypt.bat, i already reach about 70 Mh/s per device...which is not bad,right?
I wonder why you would recompile djm's work into your builds,but choose to disable support for sm3.0 devices?

[IMweasel]

You will get a lot of boo like that without the gbt check... That means somebody else found it before you. You can sometimes have more details in the wallet debug.log

I'm not sure what to look for in the debug.log...
These are my CC and Cudaminer stats. CC was up yesterday all afternoon, Cuda was up last night.
Comparing the the yay and boo rates; i think something is going wrong somewhere!
I don't have a clue what to do/try to get this working properly.

CC

[2015-06-01 19:59:16] GPU #0: n=88220
[2015-06-01 20:00:37] found => 0008dd26 26dd0800
[2015-06-01 20:00:37] GPU #0: GeForce GTX 660, 303.77 H/s
[2015-06-01 20:00:37] accepted: 0/7 (0.00%), 607.15 H/s booooo
[2015-06-01 20:00:37] job 556c9dc2 target change: 91743000000 (11.0)
[2015-06-01 20:00:37] GPU #0: start=00000000 end=000fffff range=000fffff

Cuda
[2015-06-02 08:41:02] accepted: 11/11 (100.00%), 0.38 khash/s (yay!!!)

edit:
past 6 hour tryout using CC
[2015-06-02 14:44:22] found => 0000a4f6 f6a40000
[2015-06-02 14:44:22] GPU #0: GeForce GTX 660, 269.62 H/s
[2015-06-02 14:44:23] accepted: 0/13 (0.00%), 541.01 H/s booooo
[2015-06-02 14:44:23] GPU #0: start=00000000 end=000fffff range=000fffff

and another 6 hour tryout using Cuda coming up.

[Leinaded]

Hi I'm having some problem with this miner. It doesn't run on my Windows 10 Technical Preview. Always gives the message "ccminer has stopped working" even running the miner with no arguments. Original ccminer by cbuchner1 v1.2 beta works fine in this same OS. I'm using the same config with this modded version and my graphic card is a Gigabyte GT 730 2GB DDR5 using the latest 352.84 drivers for Windows 10.

Run params: ccminer-x64-35-50-52.exe -a dmd-gr -o stratum+tcp://dmdpool.digsys.bg:3333 -u username.workername -p workerpassword

[sp_]

try with a lower intensity. like -i 10. and increase upwards if it goes well.

[Leinaded]

try with a lower intensity. like -i 10. and increase upwards if it goes well.

Thanks sp_ but it seems the problem is not intensity related. Tried that and still same crash. It doesnt even show the cmd screen. :S