Topic: [ANN] cudaMiner & ccMiner CUDA based mining applications [Windows/Linux/MacOSX] - page 271. (Read 3426936 times)

Thanks for the reply!
Thats cinda low on profit..
Anything worth mining now?

With 260 H/s you should get 0.34 XMR per Day
Maybe 20% more, as the BOT is pessimistic and accounts for 20% orphans, which is not accurate with the recent pool performance.

@tsiv: Me owes you a pizza! Let's exchange PMs on this.

Cheers,
~ Myagui

ivan, how many coins(monero) can I espect from a 750Ti ?

this commit broke mining under linux giving lot of:
reversing the commit and building again fixed it.

Does anyone have a guide on how to set up a Monero wallet on Linux?

I'm running on 8x60, haven't found anything that works better. And like you said, 8x120 is practically the same, but just a little bit slower.

8x60 gives me 261h/s per 750 ti....EVGA SC.
Was 240-243h/s
Thanks!

Instaling the new nvidia beta driver crashed ccminer no mater what algo , i had to roll back the driver to fix it. Anyone with the same problem?

...both works with kopiemtu too.
8x60 & 8x120 have nearly the same performance. 8x96 is a lot worse.
what give you the most performance on your linux rig?

thanks for the update. 

Wow, didn't see that coming. It's not every day you take a man's work, more or less turn it against him and then get paid by him. Always wondered how much I was stepping on your toes with my release, guess that answers that question. Thank you, good sir, thank you very much


Thought of that on the side, might be doable but some configs do so badly it might be TDR city all over again. I've managed values that give like 30 H/s compared to the inexplicably optimal ones that give around 280 H/s. Should probably take a poke at it anyway, at some point.


The slight relief of seeing I'm not the only one getting fsck'd by Microsoft software, priceless 


I actually compiled ccminer using 6.0 for quite some time, until I finally got fed up with editing the VC project files every time a new version came out. Nothing gained and nothing lost on going 5.5 -> 6.0 as far as I could tell.


Should be fine in theory, but if I'm not mistaken cudaMalloc requires a contiguous chunk of memory or it fails. So if there is even a tiniest allocation somewhere in the middle, the big allocation fails if there isn't a contiguous chunk of around 1.5 GB on either side of the smaller allocation. I might be wrong on this but 2 GB should be enough for 8x96. I can do 8x120 on my Linux rig with a 2 GB 750 Ti:

Aaand I'm back. Had to focus a bit on shit I've been neglecting lately, including but not limited to my actual job and sleep  

Say bye to (most of) the lag on Windows if you're willing to take a slight hit in performance. Basically I'm now splitting the heaviest part of the Cryptonight core into smaller batches with a little sleep between launching the batches. The default is still no splitting and no sleeping on Linux, on Windows it gets split into 64 batches with 100 microseconds sleep between the batches. Both values can be set on the command line per-device. I'm running it as I type this and while there is still noticeable lag, it's not too bad. The performance hit is pretty negligible on the defaults, something like 5 H/s. Well, on my system anyway.

Latest win32 binary at https://github.com/tsiv/ccminer-cryptonight/releases/tag/v0.14 and source at https://github.com/tsiv/ccminer-cryptonight as you might expect.

Obviosly this setting is too much for the 2gb ram of ti 750. But the new miner works fine with the old settings 40 blocks 8 treads.

Also getting the same errors regardless the setting:

[2014-07-05 00:09:42] 1 miner threads started, using 'cryptonight' algorithm.
[2014-07-05 00:09:42] GPU #0: GeForce GTX 750 Ti, using 96 blocks of 8 threads
[2014-07-05 00:09:42] Starting Stratum on stratum+tcp://xmr.extremehash.com:9999

[2014-07-05 00:09:42] GPU #0: FATAL: failed to allocate device memory for long s
tate

I keep getting this with random GPUs:


I'm using windows, but I'm also using the onboard GPU so exactly 0 RAM is being used on the cards.

Not sure if there is a memory backoff value in ccminer as it was in cudaminer.
(#if WIN32
device_backoff[thr_id] = 12;
#else
device_backoff[thr_id] = 2;
#endif)

The only way I don't get that problem is if I use a much lower than optimal kernel setting.

Carlo, Crashes my card in  cryptonight for me too.. 750ti, win7 64 , using stock command line, other algos working ok.
Posted from Bitcointa.lk - #arNJ7ehtPEYt5Muh

Whoops!  475 h/s was actually at -l 6x120.

Did some short-run testing (FWIW, my rig is running on 8 GB of RAM and on an i5-4670k @ 4.5 GHz.):
GeForce GTX 780 Ti 6x128 : 450 h/s
GeForce GTX 780 Ti 4x128 : 430 h/s
GeForce GTX 780 Ti 4x120 : 440 h/s

I might do some more testing at a later date for higher thread numbers, but, from what it looks like so far, other choices are still more profitable for higher-end cards (like the 780 Ti) to mine.

What algo/kernal are you trying to run?

If CryptoNight did you previously run the Reg Hack that changes the timeout of the display driver timeout?

How much memory?

How many GPU cards? 

What is the command line you are using?

Did you try a different algo just to see if that would work?

Carlo

I must be missing something mine crashes driver ver 340.43 within 3 sec's (forgot to add gpu 750ti ftw stock)

Nice.  So 4x120 on the 780 TI gives you the highest rate?  Could you do me a favor?  Please do a "cut and paste" of the specific way the card is shown. Examples:
GeForce GTX 750 Ti
GeForce GTX 660
GeForce GT 740M

Could you also try 4x128 and let me know the hash rate you get with that?  I think block sizes of multiples of 32 are closer to optimal.
Here is what I've found from testing:
GeForce GTX 750 Ti  8x96
GeForce GTX 660  7x96
GeForce GT 740M  9x64
Default for unknown cards 8x64

I'm still testing different variations but those are the best I've found so far.  If anyone has any better configs I'm all for trying it and can make it the default for those cards.

Carlo

That would be great on a 6 card rig with X11 ect.