Topic: [ANN] ccminer 2.3 - opensource - GPL (tpruvot) - page 126. (Read 500113 times)

@Epsylon3, do you have any plans for adding support for Sia algorithm? It was recently added to sgminer with stratum support: https://github.com/nicehash/sgminer/commits/master

no, there is no https http://yiimp.ccminer.org

Internet sites are accessed via http or https protocol.
It doesn't matter if one wrotes these in front...

I tried through 2 different ISP and nothing ;(, even though site opens fine through 3g on my phone

Isn't that link work for you, or what is a joke?

it's the os for sure, i tried everything and without pushing it to the maximum which would result in 180w, you can not reach 1MH, normal is 920khs, with oc 950

if you can try with win 7 you will get the same, there is another user using ccminer-vert release getting 1050 like you on linux, so this confirm that it need some fix for windows if possible

nice joke :S

Actually I'm getting more than 1 Mh/s with the lowest TDP possible (125W) on linux.
So the difference must the the OS or the drivers.

it's not, just use yiimp.ccminer.org/

Is there dns problem with yiimp ?
I can't reach it


http, https doesn't matter, same result

Amph, you know what .... i dont care

unfortunately i don't have linux, what i know is that there are guy right now reaching above 1MH with 1800 core, and i'm stuck at 900 with the same oc, something is wrong here

Thank you for your good job.
I tried to test with GTX980.

In Lyra2REv2, I think 32bit ccminer is faster than 64bit one. So please release 32bit ccminer in CUDA8.0.

so use djm version (and report me the 1080 hashrate also)

i guess it was a previous build with root... that disallow the cleanup of temporary build files

I ran ./build with Sudo, for whatever reason that fixed it.

i think neoscrypt can see a farther optimization or fix don't know, because right now you need 2GH on the core to have 1MH, which is too much, also very power hungry compared to other algo

it should in theory touch 1M without that heavy oc, and around 1200 with oc, in fact in linux(using djm34 version) it reach 1150 with the same oc

Thanks, I also tried using build.sh after making the changes to the Makefile.am but same issue.
Do you happen to know the current steps for a GeForce GTX 780 Ti ?
Git clone, cd into folder, edit makefile, ./build and also manually doing the build and make steps produce the same error.
It properly compiles however and runs, but that error spamming produces no hash.

*** ccminer 1.8 for nVidia GPUs by tpruvot@github ***
    Built with the nVidia CUDA Toolkit 7.5

  Originally based on Christian Buchner and Christian H. project
  Include some of the work of djm34, sp, tsiv and klausT.

BTC donation address: 1AJdfCpLWPNoAMDfHF1wD5y8VgKSSTHxPo (tpruvot)

[2016-07-20 19:20:28] Starting on stratum+tcp://yiimp.ccminer.org:3334
[2016-07-20 19:20:28] NVML GPU monitoring enabled.
[2016-07-20 19:20:28] 1 miner thread started, using 'lbry' algorithm.
[2016-07-20 19:20:28] Stratum difficulty set to 4
[2016-07-20 19:20:28] yiimp.ccminer.org:3334 lbry block 15115
Then it goes into the error.
[2016-07-20 19:21:40] GPU #0: scanhash_lbry:140 invalid device function
[2016-07-20 19:21:40] GPU #0: scanhash_lbry:133 invalid device function
[2016-07-20 19:21:40] GPU #0: scanhash_lbry:136 invalid device function
[2016-07-20 19:21:40] GPU #0: scanhash_lbry:140 invalid device function
[2016-07-20 19:21:40] GPU #0: scanhash_lbry:133 invalid device function
and repeats very quickly.

nvidia 840m (notebook) - 27mhs lbry ))

x86 binarie is faster by 5% versus the x64 binarie, talking about lbry

was tested on SM 3.0 and 3.5, it should work, use ./build.sh for a proper reconf

https://github.com/tpruvot/ccminer/wiki/Compatibility