[ANN] cudaMiner & ccMiner CUDA based mining applications [Windows/Linux/MacOSX]

[cbuchner1]

Why does this miner have such crappy file hosting. I have huge servers and willing to host the files for the developer/OP. PM me and its a done deal and thanks for your amazing work and program!

http://sonstiges.org/seite2.html

[1714224666]

1714224666

[bugilt]

I see Cuda 6.0 is coming soon with a few improvements will you implement cuda 6.0 into your programming when its available?

Drop-in Libraries – Automatically accelerates applications' BLAS and FFTW calculations by up to 8X by simply replacing the existing CPU libraries with the GPU-accelerated equivalents.

Would this increase mining speed?

[Valnurat]

I see alot of different setup with for an exampel Lx15 ect.

How do I find a setup for FX 4600?

I'm not sure if it is related, but when I start cudaminer the applications stops as you can see. The error is: that Windows will shutdown the app and will inform me if there is a solution for it.



I hope someone can help me?

[ArcticWolfie]

I'm trying to mine a sha256d coin, but somehow cudaminer isn't really fond of that, it starts mining with my Intel IGP and doesn't pick my GTX765M.


any idea how to fix this? If I remove all parameters it doesn't even start
I'm pretty sure it's NOT using my Nvidia GPU (because the laptop status says so). Scrypt mining works fine...

Cudaminer doesn't do SHA256, only scrypt

You realize scrypt has a part SHA256 right?

And that matter why? Cudaminer wasn't programmed to do SHA based coins, only scrypt. Don't believe me? Go read about the first 50 pages

thx, didn't know that.
the -a sha2526d function is a bit confusing though.

[cbuchner1]

Hey Mr. Andersen (dga-dave),

if you're still lurking here and if you're interested in scrypt-jane (ChaCha 20/8, Kekkac) for YaCoin and other clones.

maybe you can figure out what is wrong in my implementation of chacha_xor_core() in kepler_kernel.cu. I pretty much tried to do it in full analogy to your existing salsa_xor_core() routine. But something is amiss. The code sits in the github repo.

I would get 1.93 kHash/s in scrypt-jane hashing with the Kepler kernel on my GT 750M card, but the results won't validate (slightly more than 2GB of video RAM required).

Code:

./cudaminer -H 0 --algo=scrypt-jane -l K2x8 -m 1 -o stratum+tcp://yac.coinmine.pl:9088 -O cbuchner1.1:x

Tonight I will try to put the same code into the Titan kernel - its memory access routines are more straightforward and closer to your original code.

In comparison, this (much slower) spinlock kernel hashes just great at only 0.96 kHash/s (WARNING: nearly 4 GB of video RAM consumed)

Code:

./cudaminer -H 0 --algo=scrypt-jane -l S2x15 -o stratum+tcp://yac.coinmine.pl:9088 -O cbuchner1.1:x

NOTE: the most power efficient cards for scrypt-jane are those with few shaders (Fermi: 96 shaders like older GT 630 models),  Kepler: 384 shaders like the GT 640) , equipped with lots of GDDR3 memory at 128 bit bus bandwidth or GDDR5 RAM (2GB and more) .

[dga]

Hey Mr. Andersen (dga-dave),

if you're still lurking here and if you're interested in scrypt-jane (ChaCha 20/8, Kekkac) for YaCoin and other clones.

maybe you can figure out what is wrong in my implementation of chacha_xor_core() in kepler_kernel.cu. I pretty much tried to do it in full analogy to your existing salsa_xor_core() routine. But something is amiss. The code sits in the github repo.

I would get 1.93 kHash/s in scrypt-jane hashing with the Kepler kernel on my GT 750M card, but the results won't validate (slightly more than 2GB of video RAM required)

Code:

./cudaminer -H 0 --algo=scrypt-jane -l K2x8 -m 1 -o stratum+tcp://yac.coinmine.pl:9088 -O cbuchner1.1:x

In comparison, this (much slower) spinlock kernel hashes just great at only 0.96 kHash/s (WARNING: nearly 4 GB or video RAM consumed)

Code:

./cudaminer -H 0 --algo=scrypt-jane -l S2x15 -o stratum+tcp://yac.coinmine.pl:9088 -O cbuchner1.1:x

I'll try to take a look at it this weekend.  Semester is starting up again, so I'm going to be not very present, though.

[Schleicher]

I see Cuda 6.0 is coming soon with a few improvements will you implement cuda 6.0 into your programming when its available?

Drop-in Libraries – Automatically accelerates applications' BLAS and FFTW calculations by up to 8X by simply replacing the existing CPU libraries with the GPU-accelerated equivalents.

Would this increase mining speed?

No. cudaMiner doesn't use these libraries.

[Lacan82]

thx, didn't know that.
the -a sha2526d function is a bit confusing though.

Welcome

[RbelMonstr]

For now it might not be complete, but you can leave comments here or on the spreadsheet if you want changes or new features and all that stuff. Now I'm off to work, but will check back after.

Thank you very much for this spreadsheet.

I want to suggest Permission to edit your own entry's (if that is even possible), or just a way to re-do the survey.
I've been sloppy when i did the survey and its bothering me, so i would like to fix that.

[bathrobehero]

Thank you very much for this spreadsheet.

I want to suggest Permission to edit your own entry's (if that is even possible), or just a way to re-do the survey.
I've been sloppy when i did the survey and its bothering me, so i would like to fix that.

To be honest the whole survey/spreadhseet combo is completely new to me, but I'll read up on how to work with things.
I'm not sure if editing own entries is possible but I'll look into it later, in the meantime I'm happy to do any changes manually, just add a note or comment on the cell or just pm me here or whatever suits you.

I made a private copy of the survey/spreadheet which I can play with without ruinning the original one so tomorrow I'll my priority will be to come up with something to sort the entries.

Edit: On another note, can someone please throw me a couple of lines explaining what do I need and how can I compile the github version of cudaMiner?

[cbuchner1]

I'll try to take a look at it this weekend.  Semester is starting up again, so I'm going to be not very present, though.

No problem. Your job comes first

I think my problems arise from your primary order which has the matrix elements for the ChaCha mixing function grouped like this:

Code:

Primary Order
 0  4  8 12     thread 0 b[], bx[]
 5  9 13  1     thread 1 b[], bx[]
10 14  2  6     thread 2 b[], bx[]
15  3  7 11     thread 3 b[], bx[]

whereas I kind of expected it to be like this...

Code:

Transposed Order
 0  4  8 12
 1  5  9 13
 2  6 10 14
 3  7 11 15

I am still fiddling with the code to get it working. Your primary order (in transposed form) would be very useful for the "Operate on diagonals" step, whereas I need my transposed order for the "Operate on columns" step.

EDIT: github now has a version that can do scrypt-jane also with the Kepler kernel. But I think the big speed advantage has vanished now. Meh.

Christian

[RbelMonstr]

To be honest the whole survey/spreadhseet combo is completely new to me, but I'll read up on how to work with things.
I'm not sure if editing own entries is possible but I'll look into it later, in the meantime I'm happy to do any changes manually, just add a note or comment on the cell or just pm me here or whatever suits you.

I made a private copy of the survey/spreadheet which I can play with without ruinning the original one so tomorrow I'll my priority will be to come up with something to sort the entries.

Awesome, i'll comment on the cells as it would be the easiest way.

edit-:  I saw your to-do list, and if you want actual prises for comparrison, would a column for Model Name be usefull?

Edit: On another note, can someone please throw me a couple of lines explaining what do I need and how can I compile the github version of cudaMiner?

I've done it 6 months ago, but i can't remember how i did it. I do know it took me a lot of reading before i had it figured out.

[bathrobehero]

I've done it 6 months ago, but i can't remember how i did it. I do know it took me a lot of reading before i had it figured out.

Yeah, I've done that, and ran into like 10 different issues during compiling, mostly with missing files, and when I ironed them all out and finally ended up with a compiled version, it didn't even started.

All I want to know really is what softwares and which versions should I use.

Edit: got stuck at:

Code:

1>  Generating Code...
1>  scrypt.cpp
1>LINK : fatal error LNK1123: failure during conversion to COFF: file invalid or corrupt
1>
1>Build FAILED.

[RbelMonstr]

All I want to know really is what softwares and which versions should I use.

I assume you have checked the OP, but all needed dependencies are linked there.
I'd like to help you more but i am afraid it is above my knowledge.

[bathrobehero]

Yeah, I checked and now I found out what caused my error (wrong cvtres.exe version) and now I have a working build.
Edit: Buuuuuut, now my hashrate is about 17 Kh/s instead of 210 Kh/s regardless of settings, libcurl version or anything I could think of.  

[beachking2000]

I see Cuda 6.0 is coming soon with a few improvements will you implement cuda 6.0 into your programming when its available?

Drop-in Libraries – Automatically accelerates applications' BLAS and FFTW calculations by up to 8X by simply replacing the existing CPU libraries with the GPU-accelerated equivalents.

Would this increase mining speed?

No. cudaMiner doesn't use these libraries.

  No I know it doesn't now but in the future if it was compiled to cuda 6.0 standards I would think it would improve hash rate a bit?

[bathrobehero]

No I know it doesn't now but in the future if it was compiled to cuda 6.0 standards I would think it would improve hash rate a bit?

I think he meant cudaminer is working without using those libraries at all, as in, they are not needed.

[avemt1]

When playing with Settings of overclocking with the Asus GT 630, I get ~1024 Khash/s. Is this an issue of the software reading the card hash rate incorrectly, or did I actually find a trick to making a $45 card run like a $900 card?

[RbelMonstr]

Hai cbuchner1,

I just send my first 0.25LTC your way as a thank you for your awesome work.

-RbelMonstr

[ozahm]

I made a private copy of the survey/spreadheet which I can play with without ruinning the original one so tomorrow I'll my priority will be to come up with something to sort the entries.

To sort entries in the spreadsheet, everyone can switch the view of the spreadsheet to "list" (by clicking the 2nd entry in the 3rd pull-down menu).
With the "list"-view you can sort the data by clicking on the column names.
Although filtering with the auto filters is possible but not very feasible in the moment, cause there are not much columns with consolidated data.
I found that out myself just 5 minutes ago, perhaps you can simply include a hint in the header section of the sheet to help people like me who are not using google docs frequently.

The main problem for my personal use case is, that I can't find out fast and easy if someone else got a simular card and found out some better OC/Cudaminer setup than I have.
To support that, it would be cool if auto-filtering to the base card model could reduce the list to relevant entries.
Therefore a dropdown list in the survey form where people have to pick the card model might had helped.
But I guess we have to life now with it as it is.

Since you opt'ed for having only OC-offsets in the sheet I need some good guess (which variant of the card model is it) and google to find out what he actual clocking might be, in order to compare it with my situation. Remember some manufacturers are producing the same card model with different bios versions and clock speeds under the same marketing product name. Different cards are easier to distinguish by their part numbers like "GV-N670OC-2GD", but in an extra column for that we might not see much entries. Still  having the absolute clock values in the sheet would be much easier for my purpose.