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

[djm34]

GTX 880

http://www.techpowerup.com/199750/nvidia-geforce-gtx-880-detailed.html

3200 cores that's 5x the 750ti... (1500kHahs/s scrypt ?)
4gb vram is not a lot on an other hand... and is pretty much old gen...

230w tdp, might be a good card for none memory using coins? that is if it performs the same as 5 750Ti's

price ? Might be the issue... when we see the price of the Z...

[djm34]

It would be fantastically better if it were Maxwell cores instead of Kepler cores.

it is... but not the largest...

[I_M]

It would be fantastically better if it were Maxwell cores instead of Kepler cores.

it is... but not the largest...

You are correct, apparently I cannot read today.

All those extra cores, and only 256-bit?

[ivanlabrie]

It would be fantastically better if it were Maxwell cores instead of Kepler cores.

it is... but not the largest...

You are correct, apparently I cannot read today.

All those extra cores, and only 256-bit?

GM104, it's like the GTX 680, all over again.
But, if it mines so well, who cares?

Price it at 450 and I'm sold.

[bigjme]

Me too. Give me 5 times the performance of the 750Ti, same power, lower price. Im fine with that

[cbuchner1]

current groestlcoin performance on 3 GTX 780Ti (only mild overclock)

Code:

[2014-04-11 01:54:41] accepted: 111/111 (100.00%), 29119 khash/s (yay!!!)
[2014-04-11 01:54:46] accepted: 112/112 (100.00%), 29190 khash/s (yay!!!)
[2014-04-11 01:54:50] accepted: 113/113 (100.00%), 29117 khash/s (yay!!!)
[2014-04-11 01:54:54] accepted: 114/114 (100.00%), 29186 khash/s (yay!!!)

I am sure we can do better

Christian

[bigjme]

current groestlcoin performance on 3 GTX 780Ti (only mild overclock)

Code:

[2014-04-11 01:54:41] accepted: 111/111 (100.00%), 29119 khash/s (yay!!!)
[2014-04-11 01:54:46] accepted: 112/112 (100.00%), 29190 khash/s (yay!!!)
[2014-04-11 01:54:50] accepted: 113/113 (100.00%), 29117 khash/s (yay!!!)
[2014-04-11 01:54:54] accepted: 114/114 (100.00%), 29186 khash/s (yay!!!)

I am sure we can do better

Christian

Very nice! What are you getting with the 750Ti's?
I volenteer mine as test subjects?

On a side note, ouch to bitcoin falling to $320
Now is so not the time to sell

[ca1ne]

current groestlcoin performance on 3 GTX 780Ti (only mild overclock)

Code:

[2014-04-11 01:54:41] accepted: 111/111 (100.00%), 29119 khash/s (yay!!!)
[2014-04-11 01:54:46] accepted: 112/112 (100.00%), 29190 khash/s (yay!!!)
[2014-04-11 01:54:50] accepted: 113/113 (100.00%), 29117 khash/s (yay!!!)
[2014-04-11 01:54:54] accepted: 114/114 (100.00%), 29186 khash/s (yay!!!)

I am sure we can do better

Christian

I was wondering, wouldn't you be able to compile a module and have cudaminer/ccminer check to see if it is present, it uses your custom adaption of the algorithm, if not present uses the sph version.  Wouldn't that satisfy gpl, since without the precompiled module it would still be fully functional?  Example I have is gpl open source android kernel, that has precompiled drivers for prop hardware, etc.  The precompiled module then could confirm that your software is what is calling it.  Just a thought, instead of rewriting base code.  A simple option would to offer as you do now a precompiled version and have a md5 checksum confirmation from your module.

[cbuchner1]

I am sure we can do better

Christian

Yes we can! (Obama voice)

Code:

[2014-04-11 02:15:03] accepted: 3/3 (100.00%), 32511 khash/s (yay!!!)
[2014-04-11 02:15:07] accepted: 4/4 (100.00%), 32512 khash/s (yay!!!)
[2014-04-11 02:15:11] accepted: 5/5 (100.00%), 32493 khash/s (yay!!!)
[2014-04-11 02:15:13] accepted: 6/6 (100.00%), 32489 khash/s (yay!!!)
[2014-04-11 02:15:15] accepted: 7/7 (100.00%), 32497 khash/s (yay!!!)

[bigjme]

current groestlcoin performance on 3 GTX 780Ti (only mild overclock)

Code:

[2014-04-11 01:54:41] accepted: 111/111 (100.00%), 29119 khash/s (yay!!!)
[2014-04-11 01:54:46] accepted: 112/112 (100.00%), 29190 khash/s (yay!!!)
[2014-04-11 01:54:50] accepted: 113/113 (100.00%), 29117 khash/s (yay!!!)
[2014-04-11 01:54:54] accepted: 114/114 (100.00%), 29186 khash/s (yay!!!)

I am sure we can do better

Christian

I was wondering, wouldn't you be able to compile a module and have cudaminer/ccminer check to see if it is present, it uses your custom adaption of the algorithm, if not present uses the sph version.  Wouldn't that satisfy gpl, since without the precompiled module it would still be fully functional?  Example I have is gpl open source android kernel, that has precompiled drivers for prop hardware, etc.  The precompiled module then could confirm that your software is what is calling it.  Just a thought, instead of rewriting base code.  A simple option would to offer as you do now a precompiled version and have a md5 checksum confirmation from your module.

Hmm this could be possible. You could even have it just include the file as it does with the required libs. Yes you would need to put the algorithm into the folder with ccminer but it could work. Again the issue is that some people may take it upon themselves to give it to the amd devs.

This is a very hard thing to overcome. But if i am honest,
Cpu-miner.c is not that large of a file to rebuild. Its finding a way to build it, optimize it, and make sure it is different enough to ensure it is not classed as a clone

While i am away from university i may dig into C++ some more and see what i can get done. I learn coding quickly aslong as i can see examples and figure out what things do. I may even attempt a full rewrite of cpu-miner :p

Hopes and dreams of someone who always wants more

@christian
That was fast :p from that i expect maybe 6mh/s for a 750ti? Right now i am getting around 3mh/s per card

[cbuchner1]

Again the issue is that some people may take it upon themselves to give it to the amd devs.

well, those will have fun trying to run .cubin code on the AMD devices


750 Ti for groestlcoin:

Code:

Hash:   0000007c79c48807965126673f9806af7032ceb617abffc5711bef3a2f83c74a
Target: 000000ff00000000000000000000000000000000000000000000000000000000
[2014-04-11 02:29:55] thread 0: 19677004 hashes, 7861 khash/s
[2014-04-11 02:29:55] Total: 7861 khash/s
[2014-04-11 02:29:55] DEBUG: hash <= target
Hash:   0000001564133dbca7e069c47e64c318f9c2b2ba9da9a60eecce6223e54e7544
Target: 000000ff00000000000000000000000000000000000000000000000000000000
[2014-04-11 02:29:55] thread 0: 2971635 hashes, 7738 khash/s
[2014-04-11 02:29:55] Total: 7738 khash/s

Christian

[bigjme]

Again the issue is that some people may take it upon themselves to give it to the amd devs.

well, those will have fun trying to run .cubin code on the AMD devices

Christian

Now im very interested. Cant wait to see this finished

[cbuchner1]

Now im very interested. Cant wait to see this finished

This will turn Quark, X11, myriad into CUDA coins. This is why it's important.
Check out the 750Ti readings above.

We aren't even remotely done optimizing this....

Christian

[bigjme]

Is that on groestl?
I havent touched any of the other coins my my 750ti's
So far i have been sitting on hvc and grs

May have to have a look into which coins use those 3 algos.

[cbuchner1]

May have to have a look into which coins use those 3 algos.

Groestl is part of Quark (+AnimeCoin) and X11 and totally dominates the runtime of these...

check out this list I posted earlier:
http://www.reddit.com/r/myriadcoin/comments/21qffb/200_myriad_per_coin_help_compile_a_list_of_all/

[bigjme]

I totally forgot about that list. I read so much stuff here i forget about half of everything i read xD bookmarked now though

If those hashrates were groestl then i will see a 2.5x improvement over right now.
Its rare i actually want to learn a new coding language to understand how something works but this has me interested now. Even if i only manage to get the webserver working i will be happy.

Though i have a feeling i may have to re-code cpu-miner slowly and sort something out properly. As right now the ordering of it seems very messy and some things can be shortened down

[djm34]

current groestlcoin performance on 3 GTX 780Ti (only mild overclock)

Code:

[2014-04-11 01:54:41] accepted: 111/111 (100.00%), 29119 khash/s (yay!!!)
[2014-04-11 01:54:46] accepted: 112/112 (100.00%), 29190 khash/s (yay!!!)
[2014-04-11 01:54:50] accepted: 113/113 (100.00%), 29117 khash/s (yay!!!)
[2014-04-11 01:54:54] accepted: 114/114 (100.00%), 29186 khash/s (yay!!!)

I am sure we can do better

Christian

I was wondering, wouldn't you be able to compile a module and have cudaminer/ccminer check to see if it is present, it uses your custom adaption of the algorithm, if not present uses the sph version.  Wouldn't that satisfy gpl, since without the precompiled module it would still be fully functional?  Example I have is gpl open source android kernel, that has precompiled drivers for prop hardware, etc.  The precompiled module then could confirm that your software is what is calling it.  Just a thought, instead of rewriting base code.  A simple option would to offer as you do now a precompiled version and have a md5 checksum confirmation from your module.

Assuming, I understand correctly, you just need to write a small bash script to check with lsmod if your module is loaded then
send or not cudaminer.

[djm34]

May have to have a look into which coins use those 3 algos.

Groestl is part of Quark (+AnimeCoin) and X11 and totally dominates the runtime of these...

check out this list I posted earlier:
http://www.reddit.com/r/myriadcoin/comments/21qffb/200_myriad_per_coin_help_compile_a_list_of_all/

That seems easier to mine myriad through that thread than using a gpu... 

[coinits]

Now im very interested. Cant wait to see this finished

This will turn Quark, X11, myriad into CUDA coins. This is why it's important.
Check out the 750Ti readings above.

We aren't even remotely done optimizing this....

Christian

Suddenly you've made me look like a genius for buying 4 750TIs LOL

[sin242]

I cannot WAIT to tear into myr-gro with the latest invention from our resident mad scientist