CCminer(SP-MOD) Modded GPU kernels.

[coinut]

No problem 

Thanks for the reply

downloading beta 14 now, if I still have problems I will test out that command

[jpouza]

Not every pool support diff. adjustment via command line I think.

Here I use d=0.16 and d=0.32 as recommended by the pool, big hash => more diff.

[sp_]

Not every pool support diff. adjustment via command line I think.
Here I use d=0.16 and d=0.32 as recommended by the pool, big hash => more diff.

Yes, on 970/980 cards where the intesity is high, a higher diff is bether.

[sp_]

No problem 
Thanks for the reply
downloading beta 14 now, if I still have problems I will test out that command

Release 13 has a bug in x11 as well. No block's will be found. Solomining doesn't work. And x13/x15 got alot of HW errors. Release 14 should fix this.

[coinut]

got beta 14 up and running

Problem was still there.

changed to another pool and now showing full hashrate  

running nice and smooth.

Also tried my hand at compiling your miner sp from your Github

downloaded vs 2010/12/13 cuda 6.0 and 6.5

after trying one night i could only get your code to compile with vs 2013. It spat out a ccminer roughly the same size as the ccminer52 in your releases.
Jammed the compiled ccminer into my rig and I just got error failed to start or close to. I suspect that I compiled a version of your code compatible with gtx 970/980
not sure though.

How do you compile your code to work with 750 ti ?

  

[sp_]

The default projectfile from github only contain compute_50 and sm_50 (750ti) You need to add compute_52 and sm_52 in the cuda prefs in the project file.

[SavellM]

Little pump with r14, 980 at 1466MHz now doing 9,7MH/s X11.

Would that be like this?
ccminer52.exe -q -r 14 -R 10 -a x11

Or like:

ccminer52.exe -q -r 3 -R 14 -a x11

Or how would you do it?

Whats the diff between r and R?

[jpouza]

Little pump with r14, 980 at 1466MHz now doing 9,7MH/s X11.

Would that be like this?
ccminer52.exe -q -r 14 -R 10 -a x11

Or like:

ccminer52.exe -q -r 3 -R 14 -a x11

Or how would you do it?

Whats the diff between r and R?

Hum?
r14 I mean release14...
I use -i 20.5 for intensity

[sp_]

Would that be like this?
ccminer52.exe -q -r 14 -R 10 -a x11

Or like:
ccminer52.exe -q -r 3 -R 14 -a x11
Or how would you do it?
Whats the diff between r and R?

The R option is to ensure that the miner try to reconnect when it fails to connect. For some reason the q option give higher hashrates with less load on the cpu if you have more than 1 card in the rig. I need to look into that. The -i parameter is important, The defaults might not be good for your card. Start with -i 19 , -i 19.5 etc. untill you find the best.

[SavellM]

Ah yeah cheers guys

Was still waking up

[sp_]

How do you compile your code to work with 750 ti ?

Did you reemember to select "Release" in the build menu of visual studio 2013? The github version should compile and run on the 750ti without any changes.

[sp_]

New build from github.
-x15 fixed for 970/980 cards
-less GPU xx result for xx is not in range (hardware)
http://www.filedropper.com/release13

Can you update to tpruvot's latest API? (Like poolinfo etc). I'm using that in my monitoring tool ^^
Nice work btw! Seems like my donation earlier was worth it ^^"
Too bad my gtx760 doesn't work anymore Guess you kicked out compute 3.0

I started on the merge to tvpouet 1.5.1, but there was so many files and I got lazy. I tested the latest version on compute 3.0 (gtx 650) and it crashed. I will look into it.

Donations are welcome. More hash is comming

[SavellM]

sp_ does this ccminer support neoscrpyt?

[coinut]

How do you compile your code to work with 750 ti ?

Did you reemember to select "Release" in the build menu of visual studio 2013? The github version should compile and run on the 750ti without any changes.

yeah I did.

went to https://github.com/sp-hash/ccminer

downloaded ZIP (button RH side in the middle)

loaded ccminer.sln into vs 2013

selected release.

configuration manager set to x64

then goto "BUILD"

Then "build solution"

wait.....

after sucessfully building I then looked in

ccminer-windows\x64\release\

found ccminer.exe  8,674KB in size

copied ccminer.exe to a fresh folder and tried to run it with a .bat that I was using with one of your releases and got an error "ccminer stopped working"

This was built from your github on the 28th/Nov

not sure what I'm doing wrong

  

[sp_]

Select x86 and not x64.

[sp_]

sp_ does this ccminer support neoscrpyt?

Not yet. Noboby has ported it yet.

[Crash95]

Cuda error in file 'Z:/ccminer/x11/cuda_x11_simd512.cu' in line 630 : out of mem
ory.
Cuda error in file 'Z:/ccminer/x11/cuda_x11_simd512.cu' in line 630 : out of mem
ory.
Cuda error in file 'Z:/ccminer/x11/cuda_x11_simd512.cu' in line 630 : out of mem
ory.

[cisahasa]

sp_ does this ccminer support neoscrpyt?

Not yet. Noboby has ported it yet.

cgminer does neo/nvidia

[mmmaybe]

sp_ does this ccminer support neoscrpyt?

Not yet. Noboby has ported it yet.

cgminer does neo/nvidia

How does nvidia compare to amd on cgminer?

[scryptr]

sp_ does this ccminer support neoscrpyt?

Not yet. Noboby has ported it yet.

cgminer does neo/nvidia

How does nvidia compare to amd on cgminer?

I get 25KH/s with cgminer 3.7.8 for neoscrypt by vehre (and wolf) on Win 7 with a 550ti.  I have not been able to compile the same version on Linux and run with a 6x750ti rig.  It will compile, and run, but only produces HW errors for me.

Cgminer is not written to take advantage of Cuda architecture.  It supposedly uses OpenCL for Nvidia.  The Cuda masters have not been producing any Cuda optimized code for neoscrypt.       --scryptr